/[pcre]/code/trunk/ChangeLog
ViewVC logotype

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log


Revision 426 - (show annotations) (download)
Wed Aug 26 15:38:32 2009 UTC (4 years, 7 months ago) by ph10
File size: 168316 byte(s)
Remove restrictions on pcre_exec() partial matching.

1 ChangeLog for PCRE
2 ------------------
3
4 Version 8.00 ??-???-??
5 ----------------------
6
7 1. The table for translating pcre_compile() error codes into POSIX error codes
8 was out-of-date, and there was no check on the pcre_compile() error code
9 being within the table. This could lead to an OK return being given in
10 error.
11
12 2. Changed the call to open a subject file in pcregrep from fopen(pathname,
13 "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests
14 in a Windows environment.
15
16 3. The pcregrep --count option prints the count for each file even when it is
17 zero, as does GNU grep. However, pcregrep was also printing all files when
18 --files-with-matches was added. Now, when both options are given, it prints
19 counts only for those files that have at least one match. (GNU grep just
20 prints the file name in this circumstance, but including the count seems
21 more useful - otherwise, why use --count?) Also ensured that the
22 combination -clh just lists non-zero counts, with no names.
23
24 4. The long form of the pcregrep -F option was incorrectly implemented as
25 --fixed_strings instead of --fixed-strings. This is an incompatible change,
26 but it seems right to fix it, and I didn't think it was worth preserving
27 the old behaviour.
28
29 5. The command line items --regex=pattern and --regexp=pattern were not
30 recognized by pcregrep, which required --regex pattern or --regexp pattern
31 (with a space rather than an '='). The man page documented the '=' forms,
32 which are compatible with GNU grep; these now work.
33
34 6. No libpcreposix.pc file was created for pkg-config; there was just
35 libpcre.pc and libpcrecpp.pc. The omission has been rectified.
36
37 7. Added #ifndef SUPPORT_UCP into the pcre_ucd.c module, to reduce its size
38 when UCP support is not needed, by modifying the Python script that
39 generates it from Unicode data files. This should not matter if the module
40 is correctly used as a library, but I received one complaint about 50K of
41 unwanted data. My guess is that the person linked everything into his
42 program rather than using a library. Anyway, it does no harm.
43
44 8. A pattern such as /\x{123}{2,2}+/8 was incorrectly compiled; the trigger
45 was a minimum greater than 1 for a wide character in a possessive
46 repetition. Chaos could result.
47
48 9. The restrictions on what a pattern can contain when partial matching is
49 requested for pcre_exec() have been removed. All patterns can now be
50 partially matched by this function. In addition, if there are at least two
51 slots in the offset vector, the offsets of the first-encountered partial
52 match are set in them when PCRE_ERROR_PARTIAL is returned.
53
54
55 Version 7.9 11-Apr-09
56 ---------------------
57
58 1. When building with support for bzlib/zlib (pcregrep) and/or readline
59 (pcretest), all targets were linked against these libraries. This included
60 libpcre, libpcreposix, and libpcrecpp, even though they do not use these
61 libraries. This caused unwanted dependencies to be created. This problem
62 has been fixed, and now only pcregrep is linked with bzlib/zlib and only
63 pcretest is linked with readline.
64
65 2. The "typedef int BOOL" in pcre_internal.h that was included inside the
66 "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been
67 moved outside it again, because FALSE and TRUE are already defined in AIX,
68 but BOOL is not.
69
70 3. The pcre_config() function was treating the PCRE_MATCH_LIMIT and
71 PCRE_MATCH_LIMIT_RECURSION values as ints, when they should be long ints.
72
73 4. The pcregrep documentation said spaces were inserted as well as colons (or
74 hyphens) following file names and line numbers when outputting matching
75 lines. This is not true; no spaces are inserted. I have also clarified the
76 wording for the --colour (or --color) option.
77
78 5. In pcregrep, when --colour was used with -o, the list of matching strings
79 was not coloured; this is different to GNU grep, so I have changed it to be
80 the same.
81
82 6. When --colo(u)r was used in pcregrep, only the first matching substring in
83 each matching line was coloured. Now it goes on to look for further matches
84 of any of the test patterns, which is the same behaviour as GNU grep.
85
86 7. A pattern that could match an empty string could cause pcregrep to loop; it
87 doesn't make sense to accept an empty string match in pcregrep, so I have
88 locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
89 seems to be how GNU grep behaves.
90
91 8. The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
92 start or after a newline", because the conditional assertion was not being
93 correctly handled. The rule now is that both the assertion and what follows
94 in the first alternative must satisfy the test.
95
96 9. If auto-callout was enabled in a pattern with a conditional group whose
97 condition was an assertion, PCRE could crash during matching, both with
98 pcre_exec() and pcre_dfa_exec().
99
100 10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was
101 used for matching.
102
103 11. Unicode property support in character classes was not working for
104 characters (bytes) greater than 127 when not in UTF-8 mode.
105
106 12. Added the -M command line option to pcretest.
107
108 14. Added the non-standard REG_NOTEMPTY option to the POSIX interface.
109
110 15. Added the PCRE_NO_START_OPTIMIZE match-time option.
111
112 16. Added comments and documentation about mis-use of no_arg in the C++
113 wrapper.
114
115 17. Implemented support for UTF-8 encoding in EBCDIC environments, a patch
116 from Martin Jerabek that uses macro names for all relevant character and
117 string constants.
118
119 18. Added to pcre_internal.h two configuration checks: (a) If both EBCDIC and
120 SUPPORT_UTF8 are set, give an error; (b) If SUPPORT_UCP is set without
121 SUPPORT_UTF8, define SUPPORT_UTF8. The "configure" script handles both of
122 these, but not everybody uses configure.
123
124 19. A conditional group that had only one branch was not being correctly
125 recognized as an item that could match an empty string. This meant that an
126 enclosing group might also not be so recognized, causing infinite looping
127 (and probably a segfault) for patterns such as ^"((?(?=[a])[^"])|b)*"$
128 with the subject "ab", where knowledge that the repeated group can match
129 nothing is needed in order to break the loop.
130
131 20. If a pattern that was compiled with callouts was matched using pcre_dfa_
132 exec(), but without supplying a callout function, matching went wrong.
133
134 21. If PCRE_ERROR_MATCHLIMIT occurred during a recursion, there was a memory
135 leak if the size of the offset vector was greater than 30. When the vector
136 is smaller, the saved offsets during recursion go onto a local stack
137 vector, but for larger vectors malloc() is used. It was failing to free
138 when the recursion yielded PCRE_ERROR_MATCH_LIMIT (or any other "abnormal"
139 error, in fact).
140
141 22. There was a missing #ifdef SUPPORT_UTF8 round one of the variables in the
142 heapframe that is used only when UTF-8 support is enabled. This caused no
143 problem, but was untidy.
144
145 23. Steven Van Ingelgem's patch to CMakeLists.txt to change the name
146 CMAKE_BINARY_DIR to PROJECT_BINARY_DIR so that it works when PCRE is
147 included within another project.
148
149 24. Steven Van Ingelgem's patches to add more options to the CMake support,
150 slightly modified by me:
151
152 (a) PCRE_BUILD_TESTS can be set OFF not to build the tests, including
153 not building pcregrep.
154
155 (b) PCRE_BUILD_PCREGREP can be see OFF not to build pcregrep, but only
156 if PCRE_BUILD_TESTS is also set OFF, because the tests use pcregrep.
157
158 25. Forward references, both numeric and by name, in patterns that made use of
159 duplicate group numbers, could behave incorrectly or give incorrect errors,
160 because when scanning forward to find the reference group, PCRE was not
161 taking into account the duplicate group numbers. A pattern such as
162 ^X(?3)(a)(?|(b)|(q))(Y) is an example.
163
164 26. Changed a few more instances of "const unsigned char *" to USPTR, making
165 the feature of a custom pointer more persuasive (as requested by a user).
166
167 27. Wrapped the definitions of fileno and isatty for Windows, which appear in
168 pcretest.c, inside #ifndefs, because it seems they are sometimes already
169 pre-defined.
170
171 28. Added support for (*UTF8) at the start of a pattern.
172
173 29. Arrange for flags added by the "release type" setting in CMake to be shown
174 in the configuration summary.
175
176
177 Version 7.8 05-Sep-08
178 ---------------------
179
180 1. Replaced UCP searching code with optimized version as implemented for Ad
181 Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
182 stage table and inline lookup instead of a function, giving speed ups of 2
183 to 5 times on some simple patterns that I tested. Permission was given to
184 distribute the MultiStage2.py script that generates the tables (it's not in
185 the tarball, but is in the Subversion repository).
186
187 2. Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
188 scripts.
189
190 3. Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
191 a group with a zero qualifier. The result of the study could be incorrect,
192 or the function might crash, depending on the pattern.
193
194 4. Caseless matching was not working for non-ASCII characters in back
195 references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
196 It now works when Unicode Property Support is available.
197
198 5. In pcretest, an escape such as \x{de} in the data was always generating
199 a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
200 non-UTF-8 mode. If the value is greater than 255, it gives a warning about
201 truncation.
202
203 6. Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
204
205 7. Added two (int) casts to pcregrep when printing the difference of two
206 pointers, in case they are 64-bit values.
207
208 8. Added comments about Mac OS X stack usage to the pcrestack man page and to
209 test 2 if it fails.
210
211 9. Added PCRE_CALL_CONVENTION just before the names of all exported functions,
212 and a #define of that name to empty if it is not externally set. This is to
213 allow users of MSVC to set it if necessary.
214
215 10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
216 the convenience functions in the pcre_get.c source file.
217
218 11. An option change at the start of a pattern that had top-level alternatives
219 could cause overwriting and/or a crash. This command provoked a crash in
220 some environments:
221
222 printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
223
224 This potential security problem was recorded as CVE-2008-2371.
225
226 12. For a pattern where the match had to start at the beginning or immediately
227 after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
228 pcre_dfa_exec() could read past the end of the passed subject if there was
229 no match. To help with detecting such bugs (e.g. with valgrind), I modified
230 pcretest so that it places the subject at the end of its malloc-ed buffer.
231
232 13. The change to pcretest in 12 above threw up a couple more cases when pcre_
233 exec() might read past the end of the data buffer in UTF-8 mode.
234
235 14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
236 the data contained the byte 0x85 as part of a UTF-8 character within its
237 first line. This applied both to normal and DFA matching.
238
239 15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
240 /^[^d]*?$/8 failed to match "abc".
241
242 16. Added a missing copyright notice to pcrecpp_internal.h.
243
244 17. Make it more clear in the documentation that values returned from
245 pcre_exec() in ovector are byte offsets, not character counts.
246
247 18. Tidied a few places to stop certain compilers from issuing warnings.
248
249 19. Updated the Virtual Pascal + BCC files to compile the latest v7.7, as
250 supplied by Stefan Weber. I made a further small update for 7.8 because
251 there is a change of source arrangements: the pcre_searchfuncs.c module is
252 replaced by pcre_ucd.c.
253
254
255 Version 7.7 07-May-08
256 ---------------------
257
258 1. Applied Craig's patch to sort out a long long problem: "If we can't convert
259 a string to a long long, pretend we don't even have a long long." This is
260 done by checking for the strtoq, strtoll, and _strtoi64 functions.
261
262 2. Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
263 pre-7.6 versions, which defined a global no_arg variable instead of putting
264 it in the RE class. (See also #8 below.)
265
266 3. Remove a line of dead code, identified by coverity and reported by Nuno
267 Lopes.
268
269 4. Fixed two related pcregrep bugs involving -r with --include or --exclude:
270
271 (1) The include/exclude patterns were being applied to the whole pathnames
272 of files, instead of just to the final components.
273
274 (2) If there was more than one level of directory, the subdirectories were
275 skipped unless they satisfied the include/exclude conditions. This is
276 inconsistent with GNU grep (and could even be seen as contrary to the
277 pcregrep specification - which I improved to make it absolutely clear).
278 The action now is always to scan all levels of directory, and just
279 apply the include/exclude patterns to regular files.
280
281 5. Added the --include_dir and --exclude_dir patterns to pcregrep, and used
282 --exclude_dir in the tests to avoid scanning .svn directories.
283
284 6. Applied Craig's patch to the QuoteMeta function so that it escapes the
285 NUL character as backslash + 0 rather than backslash + NUL, because PCRE
286 doesn't support NULs in patterns.
287
288 7. Added some missing "const"s to declarations of static tables in
289 pcre_compile.c and pcre_dfa_exec.c.
290
291 8. Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
292 caused by fix #2 above. (Subsequently also a second patch to fix the
293 first patch. And a third patch - this was a messy problem.)
294
295 9. Applied Craig's patch to remove the use of push_back().
296
297 10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
298 matching function regexec().
299
300 11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
301 which, however, unlike Perl's \g{...}, are subroutine calls, not back
302 references. PCRE supports relative numbers with this syntax (I don't think
303 Oniguruma does).
304
305 12. Previously, a group with a zero repeat such as (...){0} was completely
306 omitted from the compiled regex. However, this means that if the group
307 was called as a subroutine from elsewhere in the pattern, things went wrong
308 (an internal error was given). Such groups are now left in the compiled
309 pattern, with a new opcode that causes them to be skipped at execution
310 time.
311
312 13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
313 to the way PCRE behaves:
314
315 (a) A lone ] character is dis-allowed (Perl treats it as data).
316
317 (b) A back reference to an unmatched subpattern matches an empty string
318 (Perl fails the current match path).
319
320 (c) A data ] in a character class must be notated as \] because if the
321 first data character in a class is ], it defines an empty class. (In
322 Perl it is not possible to have an empty class.) The empty class []
323 never matches; it forces failure and is equivalent to (*FAIL) or (?!).
324 The negative empty class [^] matches any one character, independently
325 of the DOTALL setting.
326
327 14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
328 non-existent subpattern following a character class starting with ']' and
329 containing () gave an internal compiling error instead of "reference to
330 non-existent subpattern". Fortunately, when the pattern did exist, the
331 compiled code was correct. (When scanning forwards to check for the
332 existencd of the subpattern, it was treating the data ']' as terminating
333 the class, so got the count wrong. When actually compiling, the reference
334 was subsequently set up correctly.)
335
336 15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
337 it was being rejected as not supported by pcre_dfa_exec(), even though
338 other assertions are supported. I have made pcre_dfa_exec() support
339 (*FAIL).
340
341 16. The implementation of 13c above involved the invention of a new opcode,
342 OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
343 cannot be changed at match time, I realized I could make a small
344 improvement to matching performance by compiling OP_ALLANY instead of
345 OP_ANY for "." when DOTALL was set, and then removing the runtime tests
346 on the OP_ANY path.
347
348 17. Compiling pcretest on Windows with readline support failed without the
349 following two fixes: (1) Make the unistd.h include conditional on
350 HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
351
352 18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
353 ncurses library to be included for pcretest when ReadLine support is
354 requested, but also to allow for it to be overridden. This patch came from
355 Daniel Bergström.
356
357 19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
358 as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
359 any errors with the current Unicode tables. Thanks to Peter Kankowski for
360 spotting this.
361
362
363 Version 7.6 28-Jan-08
364 ---------------------
365
366 1. A character class containing a very large number of characters with
367 codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
368 overflow.
369
370 2. Patch to cut out the "long long" test in pcrecpp_unittest when
371 HAVE_LONG_LONG is not defined.
372
373 3. Applied Christian Ehrlicher's patch to update the CMake build files to
374 bring them up to date and include new features. This patch includes:
375
376 - Fixed PH's badly added libz and libbz2 support.
377 - Fixed a problem with static linking.
378 - Added pcredemo. [But later removed - see 7 below.]
379 - Fixed dftables problem and added an option.
380 - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
381 HAVE_LONG_LONG.
382 - Added readline support for pcretest.
383 - Added an listing of the option settings after cmake has run.
384
385 4. A user submitted a patch to Makefile that makes it easy to create
386 "pcre.dll" under mingw when using Configure/Make. I added stuff to
387 Makefile.am that cause it to include this special target, without
388 affecting anything else. Note that the same mingw target plus all
389 the other distribution libraries and programs are now supported
390 when configuring with CMake (see 6 below) instead of with
391 Configure/Make.
392
393 5. Applied Craig's patch that moves no_arg into the RE class in the C++ code.
394 This is an attempt to solve the reported problem "pcrecpp::no_arg is not
395 exported in the Windows port". It has not yet been confirmed that the patch
396 solves the problem, but it does no harm.
397
398 6. Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
399 NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
400 with CMake, and also correct the comment about stack recursion.
401
402 7. Remove the automatic building of pcredemo from the ./configure system and
403 from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
404 of a program that users should build themselves after PCRE is installed, so
405 building it automatically is not really right. What is more, it gave
406 trouble in some build environments.
407
408 8. Further tidies to CMakeLists.txt from Sheri and Christian.
409
410
411 Version 7.5 10-Jan-08
412 ---------------------
413
414 1. Applied a patch from Craig: "This patch makes it possible to 'ignore'
415 values in parens when parsing an RE using the C++ wrapper."
416
417 2. Negative specials like \S did not work in character classes in UTF-8 mode.
418 Characters greater than 255 were excluded from the class instead of being
419 included.
420
421 3. The same bug as (2) above applied to negated POSIX classes such as
422 [:^space:].
423
424 4. PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
425 defined or documented. It seems to have been a typo for PCRE_STATIC, so
426 I have changed it.
427
428 5. The construct (?&) was not diagnosed as a syntax error (it referenced the
429 first named subpattern) and a construct such as (?&a) would reference the
430 first named subpattern whose name started with "a" (in other words, the
431 length check was missing). Both these problems are fixed. "Subpattern name
432 expected" is now given for (?&) (a zero-length name), and this patch also
433 makes it give the same error for \k'' (previously it complained that that
434 was a reference to a non-existent subpattern).
435
436 6. The erroneous patterns (?+-a) and (?-+a) give different error messages;
437 this is right because (?- can be followed by option settings as well as by
438 digits. I have, however, made the messages clearer.
439
440 7. Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
441 than the number used in the conditional) now cause a compile-time error.
442 This is actually not compatible with Perl, which accepts such patterns, but
443 treats the conditional as always being FALSE (as PCRE used to), but it
444 seems to me that giving a diagnostic is better.
445
446 8. Change "alphameric" to the more common word "alphanumeric" in comments
447 and messages.
448
449 9. Fix two occurrences of "backslash" in comments that should have been
450 "backspace".
451
452 10. Remove two redundant lines of code that can never be obeyed (their function
453 was moved elsewhere).
454
455 11. The program that makes PCRE's Unicode character property table had a bug
456 which caused it to generate incorrect table entries for sequences of
457 characters that have the same character type, but are in different scripts.
458 It amalgamated them into a single range, with the script of the first of
459 them. In other words, some characters were in the wrong script. There were
460 thirteen such cases, affecting characters in the following ranges:
461
462 U+002b0 - U+002c1
463 U+0060c - U+0060d
464 U+0061e - U+00612
465 U+0064b - U+0065e
466 U+0074d - U+0076d
467 U+01800 - U+01805
468 U+01d00 - U+01d77
469 U+01d9b - U+01dbf
470 U+0200b - U+0200f
471 U+030fc - U+030fe
472 U+03260 - U+0327f
473 U+0fb46 - U+0fbb1
474 U+10450 - U+1049d
475
476 12. The -o option (show only the matching part of a line) for pcregrep was not
477 compatible with GNU grep in that, if there was more than one match in a
478 line, it showed only the first of them. It now behaves in the same way as
479 GNU grep.
480
481 13. If the -o and -v options were combined for pcregrep, it printed a blank
482 line for every non-matching line. GNU grep prints nothing, and pcregrep now
483 does the same. The return code can be used to tell if there were any
484 non-matching lines.
485
486 14. Added --file-offsets and --line-offsets to pcregrep.
487
488 15. The pattern (?=something)(?R) was not being diagnosed as a potentially
489 infinitely looping recursion. The bug was that positive lookaheads were not
490 being skipped when checking for a possible empty match (negative lookaheads
491 and both kinds of lookbehind were skipped).
492
493 16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
494 inclusion of <windows.h> to before rather than after the definition of
495 INVALID_FILE_ATTRIBUTES (patch from David Byron).
496
497 17. Specifying a possessive quantifier with a specific limit for a Unicode
498 character property caused pcre_compile() to compile bad code, which led at
499 runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
500 are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
501 caused the error; without that there was no problem.
502
503 18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
504
505 19. Added --enable-pcretest-libreadline.
506
507 20. In pcrecpp.cc, the variable 'count' was incremented twice in
508 RE::GlobalReplace(). As a result, the number of replacements returned was
509 double what it should be. I removed one of the increments, but Craig sent a
510 later patch that removed the other one (the right fix) and added unit tests
511 that check the return values (which was not done before).
512
513 21. Several CMake things:
514
515 (1) Arranged that, when cmake is used on Unix, the libraries end up with
516 the names libpcre and libpcreposix, not just pcre and pcreposix.
517
518 (2) The above change means that pcretest and pcregrep are now correctly
519 linked with the newly-built libraries, not previously installed ones.
520
521 (3) Added PCRE_SUPPORT_LIBREADLINE, PCRE_SUPPORT_LIBZ, PCRE_SUPPORT_LIBBZ2.
522
523 22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
524 crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
525 UTF-8 newline character). The key issue is that the pattern starts .*;
526 this means that the match must be either at the beginning, or after a
527 newline. The bug was in the code for advancing after a failed match and
528 checking that the new position followed a newline. It was not taking
529 account of UTF-8 characters correctly.
530
531 23. PCRE was behaving differently from Perl in the way it recognized POSIX
532 character classes. PCRE was not treating the sequence [:...:] as a
533 character class unless the ... were all letters. Perl, however, seems to
534 allow any characters between [: and :], though of course it rejects as
535 unknown any "names" that contain non-letters, because all the known class
536 names consist only of letters. Thus, Perl gives an error for [[:1234:]],
537 for example, whereas PCRE did not - it did not recognize a POSIX character
538 class. This seemed a bit dangerous, so the code has been changed to be
539 closer to Perl. The behaviour is not identical to Perl, because PCRE will
540 diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
541 treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
542 Perl does, and where it didn't before.
543
544 24. Rewrite so as to remove the single use of %n from pcregrep because in some
545 Windows environments %n is disabled by default.
546
547
548 Version 7.4 21-Sep-07
549 ---------------------
550
551 1. Change 7.3/28 was implemented for classes by looking at the bitmap. This
552 means that a class such as [\s] counted as "explicit reference to CR or
553 LF". That isn't really right - the whole point of the change was to try to
554 help when there was an actual mention of one of the two characters. So now
555 the change happens only if \r or \n (or a literal CR or LF) character is
556 encountered.
557
558 2. The 32-bit options word was also used for 6 internal flags, but the numbers
559 of both had grown to the point where there were only 3 bits left.
560 Fortunately, there was spare space in the data structure, and so I have
561 moved the internal flags into a new 16-bit field to free up more option
562 bits.
563
564 3. The appearance of (?J) at the start of a pattern set the DUPNAMES option,
565 but did not set the internal JCHANGED flag - either of these is enough to
566 control the way the "get" function works - but the PCRE_INFO_JCHANGED
567 facility is supposed to tell if (?J) was ever used, so now (?J) at the
568 start sets both bits.
569
570 4. Added options (at build time, compile time, exec time) to change \R from
571 matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
572
573 5. doc/pcresyntax.html was missing from the distribution.
574
575 6. Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
576 compatibility, even though it is no longer used.
577
578 7. Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
579 strtoull to pcrecpp.cc to select the available functions in WIN32 when the
580 windows.h file is present (where different names are used). [This was
581 reversed later after testing - see 16 below.]
582
583 8. Changed all #include <config.h> to #include "config.h". There were also
584 some further <pcre.h> cases that I changed to "pcre.h".
585
586 9. When pcregrep was used with the --colour option, it missed the line ending
587 sequence off the lines that it output.
588
589 10. It was pointed out to me that arrays of string pointers cause lots of
590 relocations when a shared library is dynamically loaded. A technique of
591 using a single long string with a table of offsets can drastically reduce
592 these. I have refactored PCRE in four places to do this. The result is
593 dramatic:
594
595 Originally: 290
596 After changing UCP table: 187
597 After changing error message table: 43
598 After changing table of "verbs" 36
599 After changing table of Posix names 22
600
601 Thanks to the folks working on Gregex for glib for this insight.
602
603 11. --disable-stack-for-recursion caused compiling to fail unless -enable-
604 unicode-properties was also set.
605
606 12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
607
608 13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
609 checked only for CRLF.
610
611 14. Added casts to pcretest.c to avoid compiler warnings.
612
613 15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
614
615 16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
616 and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
617 entirely. This removes changes made in 7 above.
618
619 17. The CMake files have been updated, and there is now more information about
620 building with CMake in the NON-UNIX-USE document.
621
622
623 Version 7.3 28-Aug-07
624 ---------------------
625
626 1. In the rejigging of the build system that eventually resulted in 7.1, the
627 line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
628 brackets there is not right, since it causes compilers to look for an
629 installed pcre.h, not the version that is in the source that is being
630 compiled (which of course may be different). I have changed it back to:
631
632 #include "pcre.h"
633
634 I have a vague recollection that the change was concerned with compiling in
635 different directories, but in the new build system, that is taken care of
636 by the VPATH setting the Makefile.
637
638 2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
639 when the subject happened to end in the byte 0x85 (e.g. if the last
640 character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
641 characters but of course it shouldn't be taken as a newline when it is part
642 of another character. The bug was that, for an unlimited repeat of . in
643 not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
644 characters when looking for a newline.
645
646 3. A small performance improvement in the DOTALL UTF-8 mode .* case.
647
648 4. Debugging: adjusted the names of opcodes for different kinds of parentheses
649 in debug output.
650
651 5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
652 long printing in the pcrecpp unittest when running under MinGW.
653
654 6. ESC_K was left out of the EBCDIC table.
655
656 7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
657 parentheses; I made it 1000, which seemed large enough. Unfortunately, the
658 limit also applies to "virtual nesting" when a pattern is recursive, and in
659 this case 1000 isn't so big. I have been able to remove this limit at the
660 expense of backing off one optimization in certain circumstances. Normally,
661 when pcre_exec() would call its internal match() function recursively and
662 immediately return the result unconditionally, it uses a "tail recursion"
663 feature to save stack. However, when a subpattern that can match an empty
664 string has an unlimited repetition quantifier, it no longer makes this
665 optimization. That gives it a stack frame in which to save the data for
666 checking that an empty string has been matched. Previously this was taken
667 from the 1000-entry workspace that had been reserved. So now there is no
668 explicit limit, but more stack is used.
669
670 8. Applied Daniel's patches to solve problems with the import/export magic
671 syntax that is required for Windows, and which was going wrong for the
672 pcreposix and pcrecpp parts of the library. These were overlooked when this
673 problem was solved for the main library.
674
675 9. There were some crude static tests to avoid integer overflow when computing
676 the size of patterns that contain repeated groups with explicit upper
677 limits. As the maximum quantifier is 65535, the maximum group length was
678 set at 30,000 so that the product of these two numbers did not overflow a
679 32-bit integer. However, it turns out that people want to use groups that
680 are longer than 30,000 bytes (though not repeat them that many times).
681 Change 7.0/17 (the refactoring of the way the pattern size is computed) has
682 made it possible to implement the integer overflow checks in a much more
683 dynamic way, which I have now done. The artificial limitation on group
684 length has been removed - we now have only the limit on the total length of
685 the compiled pattern, which depends on the LINK_SIZE setting.
686
687 10. Fixed a bug in the documentation for get/copy named substring when
688 duplicate names are permitted. If none of the named substrings are set, the
689 functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
690 empty string.
691
692 11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
693 instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
694 because the ] is interpreted as the first data character and the
695 terminating ] is not found. PCRE has been made compatible with Perl in this
696 regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
697 cause memory overwriting.
698
699 10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
700 string has been matched (to stop an infinite loop). It was not recognizing
701 a conditional subpattern that could match an empty string if that
702 subpattern was within another subpattern. For example, it looped when
703 trying to match (((?(1)X|))*) but it was OK with ((?(1)X|)*) where the
704 condition was not nested. This bug has been fixed.
705
706 12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
707 past the start of the subject in the presence of bytes with the top bit
708 set, for example "\x8aBCD".
709
710 13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
711 (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
712
713 14. Optimized (?!) to (*FAIL).
714
715 15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
716 This restricts code points to be within the range 0 to 0x10FFFF, excluding
717 the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
718 full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
719 does: it's just the validity check that is more restrictive.
720
721 16. Inserted checks for integer overflows during escape sequence (backslash)
722 processing, and also fixed erroneous offset values for syntax errors during
723 backslash processing.
724
725 17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
726 for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
727
728 18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
729 caused an overrun.
730
731 19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
732 something other than just ASCII characters) inside a group that had an
733 unlimited repeat caused a loop at compile time (while checking to see
734 whether the group could match an empty string).
735
736 20. Debugging a pattern containing \p or \P could cause a crash. For example,
737 [\P{Any}] did so. (Error in the code for printing property names.)
738
739 21. An orphan \E inside a character class could cause a crash.
740
741 22. A repeated capturing bracket such as (A)? could cause a wild memory
742 reference during compilation.
743
744 23. There are several functions in pcre_compile() that scan along a compiled
745 expression for various reasons (e.g. to see if it's fixed length for look
746 behind). There were bugs in these functions when a repeated \p or \P was
747 present in the pattern. These operators have additional parameters compared
748 with \d, etc, and these were not being taken into account when moving along
749 the compiled data. Specifically:
750
751 (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
752 length.
753
754 (b) An item such as \pL+ within a repeated group could cause crashes or
755 loops.
756
757 (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
758 "reference to non-existent subpattern" error.
759
760 (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
761
762 24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
763 characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
764
765 25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
766
767 26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
768 character were causing crashes (broken optimization).
769
770 27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
771 \p or \P) caused a compile-time loop.
772
773 28. More problems have arisen in unanchored patterns when CRLF is a valid line
774 break. For example, the unstudied pattern [\r\n]A does not match the string
775 "\r\nA" because change 7.0/46 below moves the current point on by two
776 characters after failing to match at the start. However, the pattern \nA
777 *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
778 the same is true. There doesn't seem any very clean way out of this, but
779 what I have chosen to do makes the common cases work: PCRE now takes note
780 of whether there can be an explicit match for \r or \n anywhere in the
781 pattern, and if so, 7.0/46 no longer applies. As part of this change,
782 there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
783 pattern has explicit CR or LF references.
784
785 29. Added (*CR) etc for changing newline setting at start of pattern.
786
787
788 Version 7.2 19-Jun-07
789 ---------------------
790
791 1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
792 which is apparently normally available under Windows.
793
794 2. Re-jig the pcregrep tests with different newline settings in an attempt
795 to make them independent of the local environment's newline setting.
796
797 3. Add code to configure.ac to remove -g from the CFLAGS default settings.
798
799 4. Some of the "internals" tests were previously cut out when the link size
800 was not 2, because the output contained actual offsets. The recent new
801 "Z" feature of pcretest means that these can be cut out, making the tests
802 usable with all link sizes.
803
804 5. Implemented Stan Switzer's goto replacement for longjmp() when not using
805 stack recursion. This gives a massive performance boost under BSD, but just
806 a small improvement under Linux. However, it saves one field in the frame
807 in all cases.
808
809 6. Added more features from the forthcoming Perl 5.10:
810
811 (a) (?-n) (where n is a string of digits) is a relative subroutine or
812 recursion call. It refers to the nth most recently opened parentheses.
813
814 (b) (?+n) is also a relative subroutine call; it refers to the nth next
815 to be opened parentheses.
816
817 (c) Conditions that refer to capturing parentheses can be specified
818 relatively, for example, (?(-2)... or (?(+3)...
819
820 (d) \K resets the start of the current match so that everything before
821 is not part of it.
822
823 (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
824
825 (f) \g{name} is another synonym - part of Perl 5.10's unification of
826 reference syntax.
827
828 (g) (?| introduces a group in which the numbering of parentheses in each
829 alternative starts with the same number.
830
831 (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
832
833 7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
834 PCRE_INFO_JCHANGED.
835
836 8. A pattern such as (.*(.)?)* caused pcre_exec() to fail by either not
837 terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
838 for detecting groups that can match an empty string.
839
840 9. A pattern with a very large number of alternatives (more than several
841 hundred) was running out of internal workspace during the pre-compile
842 phase, where pcre_compile() figures out how much memory will be needed. A
843 bit of new cunning has reduced the workspace needed for groups with
844 alternatives. The 1000-alternative test pattern now uses 12 bytes of
845 workspace instead of running out of the 4096 that are available.
846
847 10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
848
849 11. Applied patch from Google to remove an optimization that didn't quite work.
850 The report of the bug said:
851
852 pcrecpp::RE("a*").FullMatch("aaa") matches, while
853 pcrecpp::RE("a*?").FullMatch("aaa") does not, and
854 pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
855
856 12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
857 it matched the wrong number of bytes.
858
859
860 Version 7.1 24-Apr-07
861 ---------------------
862
863 1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
864 that is more "standard", making use of automake and other Autotools. There
865 is some re-arrangement of the files and adjustment of comments consequent
866 on this.
867
868 2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
869 for recursive directory scanning broke on some systems because the files
870 are not scanned in any specific order and on different systems the order
871 was different. A call to "sort" has been inserted into RunGrepTest for the
872 approprate test as a short-term fix. In the longer term there may be an
873 alternative.
874
875 3. I had an email from Eric Raymond about problems translating some of PCRE's
876 man pages to HTML (despite the fact that I distribute HTML pages, some
877 people do their own conversions for various reasons). The problems
878 concerned the use of low-level troff macros .br and .in. I have therefore
879 removed all such uses from the man pages (some were redundant, some could
880 be replaced by .nf/.fi pairs). The 132html script that I use to generate
881 HTML has been updated to handle .nf/.fi and to complain if it encounters
882 .br or .in.
883
884 4. Updated comments in configure.ac that get placed in config.h.in and also
885 arranged for config.h to be included in the distribution, with the name
886 config.h.generic, for the benefit of those who have to compile without
887 Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
888
889 5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
890 Weber: (1) pcre_internal.h was missing some function renames; (2) updated
891 makevp.bat for the current PCRE, using the additional files
892 makevp_c.txt, makevp_l.txt, and pcregexp.pas.
893
894 6. A Windows user reported a minor discrepancy with test 2, which turned out
895 to be caused by a trailing space on an input line that had got lost in his
896 copy. The trailing space was an accident, so I've just removed it.
897
898 7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
899 that is needed.
900
901 8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
902 as "const" (a) because they are and (b) because it helps the PHP
903 maintainers who have recently made a script to detect big data structures
904 in the php code that should be moved to the .rodata section. I remembered
905 to update Builducptable as well, so it won't revert if ucptable.h is ever
906 re-created.
907
908 9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
909 pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
910 order to be able to cut out the UTF-8 tables in the latter when UTF-8
911 support is not required. This saves 1.5-2K of code, which is important in
912 some applications.
913
914 Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
915 so as not to refer to the tables, even though these functions will never be
916 called when UTF-8 support is disabled. Otherwise there are problems with a
917 shared library.
918
919 10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
920
921 (a) It was defining its arguments as char * instead of void *.
922
923 (b) It was assuming that all moves were upwards in memory; this was true
924 a long time ago when I wrote it, but is no longer the case.
925
926 The emulated memove() is provided for those environments that have neither
927 memmove() nor bcopy(). I didn't think anyone used it these days, but that
928 is clearly not the case, as these two bugs were recently reported.
929
930 11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
931 and Detrail to create the HTML documentation, the .txt form of the man
932 pages, and it removes trailing spaces from listed files. It also creates
933 pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
934 case, it wraps all the #defines with #ifndefs. This script should be run
935 before "make dist".
936
937 12. Fixed two fairly obscure bugs concerned with quantified caseless matching
938 with Unicode property support.
939
940 (a) For a maximizing quantifier, if the two different cases of the
941 character were of different lengths in their UTF-8 codings (there are
942 some cases like this - I found 11), and the matching function had to
943 back up over a mixture of the two cases, it incorrectly assumed they
944 were both the same length.
945
946 (b) When PCRE was configured to use the heap rather than the stack for
947 recursion during matching, it was not correctly preserving the data for
948 the other case of a UTF-8 character when checking ahead for a match
949 while processing a minimizing repeat. If the check also involved
950 matching a wide character, but failed, corruption could cause an
951 erroneous result when trying to check for a repeat of the original
952 character.
953
954 13. Some tidying changes to the testing mechanism:
955
956 (a) The RunTest script now detects the internal link size and whether there
957 is UTF-8 and UCP support by running ./pcretest -C instead of relying on
958 values substituted by "configure". (The RunGrepTest script already did
959 this for UTF-8.) The configure.ac script no longer substitutes the
960 relevant variables.
961
962 (b) The debugging options /B and /D in pcretest show the compiled bytecode
963 with length and offset values. This means that the output is different
964 for different internal link sizes. Test 2 is skipped for link sizes
965 other than 2 because of this, bypassing the problem. Unfortunately,
966 there was also a test in test 3 (the locale tests) that used /B and
967 failed for link sizes other than 2. Rather than cut the whole test out,
968 I have added a new /Z option to pcretest that replaces the length and
969 offset values with spaces. This is now used to make test 3 independent
970 of link size. (Test 2 will be tidied up later.)
971
972 14. If erroroffset was passed as NULL to pcre_compile, it provoked a
973 segmentation fault instead of returning the appropriate error message.
974
975 15. In multiline mode when the newline sequence was set to "any", the pattern
976 ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
977 This doesn't seem right; it now treats the CRLF combination as the line
978 ending, and so does not match in that case. It's only a pattern such as ^$
979 that would hit this one: something like ^ABC$ would have failed after \r
980 and then tried again after \r\n.
981
982 16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
983 in an attempt to make files that differ only in their line terminators
984 compare equal. This works on Linux.
985
986 17. Under certain error circumstances pcregrep might try to free random memory
987 as it exited. This is now fixed, thanks to valgrind.
988
989 19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
990 "abc\r\n\r\n", it found an unwanted second match after the second \r. This
991 was because its rules for how to advance for /g after matching an empty
992 string at the end of a line did not allow for this case. They now check for
993 it specially.
994
995 20. pcretest is supposed to handle patterns and data of any length, by
996 extending its buffers when necessary. It was getting this wrong when the
997 buffer for a data line had to be extended.
998
999 21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
1000 CRLF as a newline sequence.
1001
1002 22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
1003 out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
1004 I have nevertheless tidied it up.
1005
1006 23. Added some casts to kill warnings from HP-UX ia64 compiler.
1007
1008 24. Added a man page for pcre-config.
1009
1010
1011 Version 7.0 19-Dec-06
1012 ---------------------
1013
1014 1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
1015 moving to gcc 4.1.1.
1016
1017 2. The -S option for pcretest uses setrlimit(); I had omitted to #include
1018 sys/time.h, which is documented as needed for this function. It doesn't
1019 seem to matter on Linux, but it showed up on some releases of OS X.
1020
1021 3. It seems that there are systems where bytes whose values are greater than
1022 127 match isprint() in the "C" locale. The "C" locale should be the
1023 default when a C program starts up. In most systems, only ASCII printing
1024 characters match isprint(). This difference caused the output from pcretest
1025 to vary, making some of the tests fail. I have changed pcretest so that:
1026
1027 (a) When it is outputting text in the compiled version of a pattern, bytes
1028 other than 32-126 are always shown as hex escapes.
1029
1030 (b) When it is outputting text that is a matched part of a subject string,
1031 it does the same, unless a different locale has been set for the match
1032 (using the /L modifier). In this case, it uses isprint() to decide.
1033
1034 4. Fixed a major bug that caused incorrect computation of the amount of memory
1035 required for a compiled pattern when options that changed within the
1036 pattern affected the logic of the preliminary scan that determines the
1037 length. The relevant options are -x, and -i in UTF-8 mode. The result was
1038 that the computed length was too small. The symptoms of this bug were
1039 either the PCRE error "internal error: code overflow" from pcre_compile(),
1040 or a glibc crash with a message such as "pcretest: free(): invalid next
1041 size (fast)". Examples of patterns that provoked this bug (shown in
1042 pcretest format) are:
1043
1044 /(?-x: )/x
1045 /(?x)(?-x: \s*#\s*)/
1046 /((?i)[\x{c0}])/8
1047 /(?i:[\x{c0}])/8
1048
1049 HOWEVER: Change 17 below makes this fix obsolete as the memory computation
1050 is now done differently.
1051
1052 5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
1053 wrapper classes; (b) implement a new function in the C++ scanner that is
1054 more efficient than the old way of doing things because it avoids levels of
1055 recursion in the regex matching; (c) add a paragraph to the documentation
1056 for the FullMatch() function.
1057
1058 6. The escape sequence \n was being treated as whatever was defined as
1059 "newline". Not only was this contrary to the documentation, which states
1060 that \n is character 10 (hex 0A), but it also went horribly wrong when
1061 "newline" was defined as CRLF. This has been fixed.
1062
1063 7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
1064 was being set to -1 for the "end of line" case (supposedly a value that no
1065 character can have). Though this value is never used (the check for end of
1066 line is "zero bytes in current character"), it caused compiler complaints.
1067 I've changed it to 0xffffffff.
1068
1069 8. In pcre_version.c, the version string was being built by a sequence of
1070 C macros that, in the event of PCRE_PRERELEASE being defined as an empty
1071 string (as it is for production releases) called a macro with an empty
1072 argument. The C standard says the result of this is undefined. The gcc
1073 compiler treats it as an empty string (which was what was wanted) but it is
1074 reported that Visual C gives an error. The source has been hacked around to
1075 avoid this problem.
1076
1077 9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
1078 builds of pcretest, and changed the call to _setmode() to use _O_BINARY
1079 instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
1080 of them did).
1081
1082 10. Originally, pcretest opened its input and output without "b"; then I was
1083 told that "b" was needed in some environments, so it was added for release
1084 5.0 to both the input and output. (It makes no difference on Unix-like
1085 systems.) Later I was told that it is wrong for the input on Windows. I've
1086 now abstracted the modes into two macros, to make it easier to fiddle with
1087 them, and removed "b" from the input mode under Windows.
1088
1089 11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
1090
1091 12. Added -help and --help to pcretest as an official way of being reminded
1092 of the options.
1093
1094 13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
1095 and pcrecpp.cc because they annoy compilers at high warning levels.
1096
1097 14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
1098
1099 15. Fixed an occurrence of == in configure.ac that should have been = (shell
1100 scripts are not C programs :-) and which was not noticed because it works
1101 on Linux.
1102
1103 16. pcretest is supposed to handle any length of pattern and data line (as one
1104 line or as a continued sequence of lines) by extending its input buffer if
1105 necessary. This feature was broken for very long pattern lines, leading to
1106 a string of junk being passed to pcre_compile() if the pattern was longer
1107 than about 50K.
1108
1109 17. I have done a major re-factoring of the way pcre_compile() computes the
1110 amount of memory needed for a compiled pattern. Previously, there was code
1111 that made a preliminary scan of the pattern in order to do this. That was
1112 OK when PCRE was new, but as the facilities have expanded, it has become
1113 harder and harder to keep it in step with the real compile phase, and there
1114 have been a number of bugs (see for example, 4 above). I have now found a
1115 cunning way of running the real compile function in a "fake" mode that
1116 enables it to compute how much memory it would need, while actually only
1117 ever using a few hundred bytes of working memory and without too many
1118 tests of the mode. This should make future maintenance and development
1119 easier. A side effect of this work is that the limit of 200 on the nesting
1120 depth of parentheses has been removed (though this was never a serious
1121 limitation, I suspect). However, there is a downside: pcre_compile() now
1122 runs more slowly than before (30% or more, depending on the pattern). I
1123 hope this isn't a big issue. There is no effect on runtime performance.
1124
1125 18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
1126 newline (only possible for the last line of a file) and it was a
1127 pattern that set a locale (followed by /Lsomething), pcretest crashed.
1128
1129 19. Added additional timing features to pcretest. (1) The -tm option now times
1130 matching only, not compiling. (2) Both -t and -tm can be followed, as a
1131 separate command line item, by a number that specifies the number of
1132 repeats to use when timing. The default is 50000; this gives better
1133 precision, but takes uncomfortably long for very large patterns.
1134
1135 20. Extended pcre_study() to be more clever in cases where a branch of a
1136 subpattern has no definite first character. For example, (a*|b*)[cd] would
1137 previously give no result from pcre_study(). Now it recognizes that the
1138 first character must be a, b, c, or d.
1139
1140 21. There was an incorrect error "recursive call could loop indefinitely" if
1141 a subpattern (or the entire pattern) that was being tested for matching an
1142 empty string contained only one non-empty item after a nested subpattern.
1143 For example, the pattern (?>\x{100}*)\d(?R) provoked this error
1144 incorrectly, because the \d was being skipped in the check.
1145
1146 22. The pcretest program now has a new pattern option /B and a command line
1147 option -b, which is equivalent to adding /B to every pattern. This causes
1148 it to show the compiled bytecode, without the additional information that
1149 -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
1150 is the same as /B/I).
1151
1152 23. A new optimization is now able automatically to treat some sequences such
1153 as a*b as a*+b. More specifically, if something simple (such as a character
1154 or a simple class like \d) has an unlimited quantifier, and is followed by
1155 something that cannot possibly match the quantified thing, the quantifier
1156 is automatically "possessified".
1157
1158 24. A recursive reference to a subpattern whose number was greater than 39
1159 went wrong under certain circumstances in UTF-8 mode. This bug could also
1160 have affected the operation of pcre_study().
1161
1162 25. Realized that a little bit of performance could be had by replacing
1163 (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
1164
1165 26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
1166
1167 27. Possessive quantifiers such as a++ were previously implemented by turning
1168 them into atomic groups such as ($>a+). Now they have their own opcodes,
1169 which improves performance. This includes the automatically created ones
1170 from 23 above.
1171
1172 28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
1173 lookahead was broken if it was not anchored. PCRE was mistakenly expecting
1174 the first matched character to be a colon. This applied both to named and
1175 numbered groups.
1176
1177 29. The ucpinternal.h header file was missing its idempotency #ifdef.
1178
1179 30. I was sent a "project" file called libpcre.a.dev which I understand makes
1180 building PCRE on Windows easier, so I have included it in the distribution.
1181
1182 31. There is now a check in pcretest against a ridiculously large number being
1183 returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
1184 loop, the loop is abandoned.
1185
1186 32. Forward references to subpatterns in conditions such as (?(2)...) where
1187 subpattern 2 is defined later cause pcre_compile() to search forwards in
1188 the pattern for the relevant set of parentheses. This search went wrong
1189 when there were unescaped parentheses in a character class, parentheses
1190 escaped with \Q...\E, or parentheses in a #-comment in /x mode.
1191
1192 33. "Subroutine" calls and backreferences were previously restricted to
1193 referencing subpatterns earlier in the regex. This restriction has now
1194 been removed.
1195
1196 34. Added a number of extra features that are going to be in Perl 5.10. On the
1197 whole, these are just syntactic alternatives for features that PCRE had
1198 previously implemented using the Python syntax or my own invention. The
1199 other formats are all retained for compatibility.
1200
1201 (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
1202 as (?P<name>...). The new forms, as well as being in Perl 5.10, are
1203 also .NET compatible.
1204
1205 (b) A recursion or subroutine call to a named group can now be defined as
1206 (?&name) as well as (?P>name).
1207
1208 (c) A backreference to a named group can now be defined as \k<name> or
1209 \k'name' as well as (?P=name). The new forms, as well as being in Perl
1210 5.10, are also .NET compatible.
1211
1212 (d) A conditional reference to a named group can now use the syntax
1213 (?(<name>) or (?('name') as well as (?(name).
1214
1215 (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
1216 groups (named and numbered) that are never evaluated inline, but can be
1217 called as "subroutines" from elsewhere. In effect, the DEFINE condition
1218 is always false. There may be only one alternative in such a group.
1219
1220 (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
1221 as the simple (?(R). The condition is true only if the most recent
1222 recursion is that of the given number or name. It does not search out
1223 through the entire recursion stack.
1224
1225 (g) The escape \gN or \g{N} has been added, where N is a positive or
1226 negative number, specifying an absolute or relative reference.
1227
1228 35. Tidied to get rid of some further signed/unsigned compiler warnings and
1229 some "unreachable code" warnings.
1230
1231 36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
1232 things, this adds five new scripts.
1233
1234 37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
1235 There were also incompatibilities regarding the handling of \Q..\E inside
1236 character classes, for example with patterns like [\Qa\E-\Qz\E] where the
1237 hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
1238
1239 38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
1240 matches an empty string, and forcibly breaks the loop. There were bugs in
1241 this code in non-simple cases. For a pattern such as ^(a()*)* matched
1242 against aaaa the result was just "a" rather than "aaaa", for example. Two
1243 separate and independent bugs (that affected different cases) have been
1244 fixed.
1245
1246 39. Refactored the code to abolish the use of different opcodes for small
1247 capturing bracket numbers. This is a tidy that I avoided doing when I
1248 removed the limit on the number of capturing brackets for 3.5 back in 2001.
1249 The new approach is not only tidier, it makes it possible to reduce the
1250 memory needed to fix the previous bug (38).
1251
1252 40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
1253 sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
1254 processing dot, circumflex, or dollar metacharacters, or #-comments in /x
1255 mode.
1256
1257 41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
1258 report.
1259
1260 42. Applied patch, originally from Ari Pollak, modified by Google, to allow
1261 copy construction and assignment in the C++ wrapper.
1262
1263 43. Updated pcregrep to support "--newline=any". In the process, I fixed a
1264 couple of bugs that could have given wrong results in the "--newline=crlf"
1265 case.
1266
1267 44. Added a number of casts and did some reorganization of signed/unsigned int
1268 variables following suggestions from Dair Grant. Also renamed the variable
1269 "this" as "item" because it is a C++ keyword.
1270
1271 45. Arranged for dftables to add
1272
1273 #include "pcre_internal.h"
1274
1275 to pcre_chartables.c because without it, gcc 4.x may remove the array
1276 definition from the final binary if PCRE is built into a static library and
1277 dead code stripping is activated.
1278
1279 46. For an unanchored pattern, if a match attempt fails at the start of a
1280 newline sequence, and the newline setting is CRLF or ANY, and the next two
1281 characters are CRLF, advance by two characters instead of one.
1282
1283
1284 Version 6.7 04-Jul-06
1285 ---------------------
1286
1287 1. In order to handle tests when input lines are enormously long, pcretest has
1288 been re-factored so that it automatically extends its buffers when
1289 necessary. The code is crude, but this _is_ just a test program. The
1290 default size has been increased from 32K to 50K.
1291
1292 2. The code in pcre_study() was using the value of the re argument before
1293 testing it for NULL. (Of course, in any sensible call of the function, it
1294 won't be NULL.)
1295
1296 3. The memmove() emulation function in pcre_internal.h, which is used on
1297 systems that lack both memmove() and bcopy() - that is, hardly ever -
1298 was missing a "static" storage class specifier.
1299
1300 4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
1301 containing an extended class (one that cannot be represented by a bitmap
1302 because it contains high-valued characters or Unicode property items, e.g.
1303 [\pZ]). Almost always one would set UTF-8 mode when processing such a
1304 pattern, but PCRE should not loop if you do not (it no longer does).
1305 [Detail: two cases were found: (a) a repeated subpattern containing an
1306 extended class; (b) a recursive reference to a subpattern that followed a
1307 previous extended class. It wasn't skipping over the extended class
1308 correctly when UTF-8 mode was not set.]
1309
1310 5. A negated single-character class was not being recognized as fixed-length
1311 in lookbehind assertions such as (?<=[^f]), leading to an incorrect
1312 compile error "lookbehind assertion is not fixed length".
1313
1314 6. The RunPerlTest auxiliary script was showing an unexpected difference
1315 between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
1316 write a Perl script that can interpret lines of an input file either as
1317 byte characters or as UTF-8, which is what "perltest" was being required to
1318 do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
1319 can't do is switch easily at run time between having the "use utf8;" pragma
1320 or not. In the end, I fudged it by using the RunPerlTest script to insert
1321 "use utf8;" explicitly for the UTF-8 tests.
1322
1323 7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
1324 the end of the subject string, contrary to the documentation and to what
1325 Perl does. This was true of both matching functions. Now it matches only at
1326 the start of the subject and immediately after *internal* newlines.
1327
1328 8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
1329 a pointer to an int instead of a pointer to an unsigned long int. This
1330 caused problems on 64-bit systems.
1331
1332 9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
1333 instance of the 'standard' template library not being so standard".
1334
1335 10. There was no check on the number of named subpatterns nor the maximum
1336 length of a subpattern name. The product of these values is used to compute
1337 the size of the memory block for a compiled pattern. By supplying a very
1338 long subpattern name and a large number of named subpatterns, the size
1339 computation could be caused to overflow. This is now prevented by limiting
1340 the length of names to 32 characters, and the number of named subpatterns
1341 to 10,000.
1342
1343 11. Subpatterns that are repeated with specific counts have to be replicated in
1344 the compiled pattern. The size of memory for this was computed from the
1345 length of the subpattern and the repeat count. The latter is limited to
1346 65535, but there was no limit on the former, meaning that integer overflow
1347 could in principle occur. The compiled length of a repeated subpattern is
1348 now limited to 30,000 bytes in order to prevent this.
1349
1350 12. Added the optional facility to have named substrings with the same name.
1351
1352 13. Added the ability to use a named substring as a condition, using the
1353 Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
1354 are numbers (not recommended). Forward references are permitted.
1355
1356 14. Added forward references in named backreferences (if you see what I mean).
1357
1358 15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
1359 pattern could run off the end of the subject. For example, the pattern
1360 "(?s)(.{1,5})"8 did this with the subject "ab".
1361
1362 16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
1363 PCRE_CASELESS was set when matching characters that were quantified with ?
1364 or *.
1365
1366 17. A character class other than a single negated character that had a minimum
1367 but no maximum quantifier - for example [ab]{6,} - was not handled
1368 correctly by pce_dfa_exec(). It would match only one character.
1369
1370 18. A valid (though odd) pattern that looked like a POSIX character
1371 class but used an invalid character after [ (for example [[,abc,]]) caused
1372 pcre_compile() to give the error "Failed: internal error: code overflow" or
1373 in some cases to crash with a glibc free() error. This could even happen if
1374 the pattern terminated after [[ but there just happened to be a sequence of
1375 letters, a binary zero, and a closing ] in the memory that followed.
1376
1377 19. Perl's treatment of octal escapes in the range \400 to \777 has changed
1378 over the years. Originally (before any Unicode support), just the bottom 8
1379 bits were taken. Thus, for example, \500 really meant \100. Nowadays the
1380 output from "man perlunicode" includes this:
1381
1382 The regular expression compiler produces polymorphic opcodes. That
1383 is, the pattern adapts to the data and automatically switches to
1384 the Unicode character scheme when presented with Unicode data--or
1385 instead uses a traditional byte scheme when presented with byte
1386 data.
1387
1388 Sadly, a wide octal escape does not cause a switch, and in a string with
1389 no other multibyte characters, these octal escapes are treated as before.
1390 Thus, in Perl, the pattern /\500/ actually matches \100 but the pattern
1391 /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
1392 Unicode string.
1393
1394 I have not perpetrated such confusion in PCRE. Up till now, it took just
1395 the bottom 8 bits, as in old Perl. I have now made octal escapes with
1396 values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
1397 translate to the appropriate multibyte character.
1398
1399 29. Applied some refactoring to reduce the number of warnings from Microsoft
1400 and Borland compilers. This has included removing the fudge introduced
1401 seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
1402 a warning about an unused variable.
1403
1404 21. PCRE has not included VT (character 0x0b) in the set of whitespace
1405 characters since release 4.0, because Perl (from release 5.004) does not.
1406 [Or at least, is documented not to: some releases seem to be in conflict
1407 with the documentation.] However, when a pattern was studied with
1408 pcre_study() and all its branches started with \s, PCRE still included VT
1409 as a possible starting character. Of course, this did no harm; it just
1410 caused an unnecessary match attempt.
1411
1412 22. Removed a now-redundant internal flag bit that recorded the fact that case
1413 dependency changed within the pattern. This was once needed for "required
1414 byte" processing, but is no longer used. This recovers a now-scarce options
1415 bit. Also moved the least significant internal flag bit to the most-
1416 significant bit of the word, which was not previously used (hangover from
1417 the days when it was an int rather than a uint) to free up another bit for
1418 the future.
1419
1420 23. Added support for CRLF line endings as well as CR and LF. As well as the
1421 default being selectable at build time, it can now be changed at runtime
1422 via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
1423 specify that it is scanning data with non-default line endings.
1424
1425 24. Changed the definition of CXXLINK to make it agree with the definition of
1426 LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
1427
1428 25. Applied Ian Taylor's patches to avoid using another stack frame for tail
1429 recursions. This makes a big different to stack usage for some patterns.
1430
1431 26. If a subpattern containing a named recursion or subroutine reference such
1432 as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
1433 the space required for the compiled pattern went wrong and gave too small a
1434 value. Depending on the environment, this could lead to "Failed: internal
1435 error: code overflow at offset 49" or "glibc detected double free or
1436 corruption" errors.
1437
1438 27. Applied patches from Google (a) to support the new newline modes and (b) to
1439 advance over multibyte UTF-8 characters in GlobalReplace.
1440
1441 28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
1442 difference for some implementation of PCRE in some Windows version.
1443
1444 29. Added some extra testing facilities to pcretest:
1445
1446 \q<number> in a data line sets the "match limit" value
1447 \Q<number> in a data line sets the "match recursion limt" value
1448 -S <number> sets the stack size, where <number> is in megabytes
1449
1450 The -S option isn't available for Windows.
1451
1452
1453 Version 6.6 06-Feb-06
1454 ---------------------
1455
1456 1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
1457 in pcreposix.h. I have copied the definition from pcre.h.
1458
1459 2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
1460 because pcre.h is no longer a built file.
1461
1462 3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
1463 not normally included in the compiled code.
1464
1465
1466 Version 6.5 01-Feb-06
1467 ---------------------
1468
1469 1. When using the partial match feature with pcre_dfa_exec(), it was not
1470 anchoring the second and subsequent partial matches at the new starting
1471 point. This could lead to incorrect results. For example, with the pattern
1472 /1234/, partially matching against "123" and then "a4" gave a match.
1473
1474 2. Changes to pcregrep:
1475
1476 (a) All non-match returns from pcre_exec() were being treated as failures
1477 to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
1478 error message is output. Some extra information is given for the
1479 PCRE_ERROR_MATCHLIMIT and PCRE_ERROR_RECURSIONLIMIT errors, which are
1480 probably the only errors that are likely to be caused by users (by
1481 specifying a regex that has nested indefinite repeats, for instance).
1482 If there are more than 20 of these errors, pcregrep is abandoned.
1483
1484 (b) A binary zero was treated as data while matching, but terminated the
1485 output line if it was written out. This has been fixed: binary zeroes
1486 are now no different to any other data bytes.
1487
1488 (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
1489 used to set a locale for matching. The --locale=xxxx long option has
1490 been added (no short equivalent) to specify a locale explicitly on the
1491 pcregrep command, overriding the environment variables.
1492
1493 (d) When -B was used with -n, some line numbers in the output were one less
1494 than they should have been.
1495
1496 (e) Added the -o (--only-matching) option.
1497
1498 (f) If -A or -C was used with -c (count only), some lines of context were
1499 accidentally printed for the final match.
1500
1501 (g) Added the -H (--with-filename) option.
1502
1503 (h) The combination of options -rh failed to suppress file names for files
1504 that were found from directory arguments.
1505
1506 (i) Added the -D (--devices) and -d (--directories) options.
1507
1508 (j) Added the -F (--fixed-strings) option.
1509
1510 (k) Allow "-" to be used as a file name for -f as well as for a data file.
1511
1512 (l) Added the --colo(u)r option.
1513
1514 (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
1515 is not present by default.
1516
1517 3. A nasty bug was discovered in the handling of recursive patterns, that is,
1518 items such as (?R) or (?1), when the recursion could match a number of
1519 alternatives. If it matched one of the alternatives, but subsequently,
1520 outside the recursion, there was a failure, the code tried to back up into
1521 the recursion. However, because of the way PCRE is implemented, this is not
1522 possible, and the result was an incorrect result from the match.
1523
1524 In order to prevent this happening, the specification of recursion has
1525 been changed so that all such subpatterns are automatically treated as
1526 atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
1527
1528 4. I had overlooked the fact that, in some locales, there are characters for
1529 which isalpha() is true but neither isupper() nor islower() are true. In
1530 the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
1531 and ordfeminine) are like this. This affected the treatment of \w and \W
1532 when they appeared in character classes, but not when they appeared outside
1533 a character class. The bit map for "word" characters is now created
1534 separately from the results of isalnum() instead of just taking it from the
1535 upper, lower, and digit maps. (Plus the underscore character, of course.)
1536
1537 5. The above bug also affected the handling of POSIX character classes such as
1538 [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
1539 permanent tables. Instead, the bit maps for such a class were previously
1540 created as the appropriate unions of the upper, lower, and digit bitmaps.
1541 Now they are created by subtraction from the [[:word:]] class, which has
1542 its own bitmap.
1543
1544 6. The [[:blank:]] character class matches horizontal, but not vertical space.
1545 It is created by subtracting the vertical space characters (\x09, \x0a,
1546 \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
1547 subtraction was done in the overall bitmap for a character class, meaning
1548 that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
1549 be recognized. This bug has been fixed.
1550
1551 7. Patches from the folks at Google:
1552
1553 (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
1554 real life, but is still worth protecting against".
1555
1556 (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
1557 regular expressions".
1558
1559 (c) pcre_scanner.cc: avoid use of std::count() because not all systems
1560 have it.
1561
1562 (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
1563 "configure" and the latter not, in order to fix a problem somebody had
1564 with compiling the Arg class on HP-UX.
1565
1566 (e) Improve the error-handling of the C++ wrapper a little bit.
1567
1568 (f) New tests for checking recursion limiting.
1569
1570 8. The pcre_memmove() function, which is used only if the environment does not
1571 have a standard memmove() function (and is therefore rarely compiled),
1572 contained two bugs: (a) use of int instead of size_t, and (b) it was not
1573 returning a result (though PCRE never actually uses the result).
1574
1575 9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
1576 large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
1577 returned instead of calling malloc() with an overflowing number that would
1578 most likely cause subsequent chaos.
1579
1580 10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
1581
1582 11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
1583 with this option is matched, the nmatch and pmatch options of regexec() are
1584 ignored.
1585
1586 12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
1587 provided in case anyone wants to the the POSIX interface with UTF-8
1588 strings.
1589
1590 13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
1591 C++ linking (needed for some HP-UX environments).
1592
1593 14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
1594 (unused parameter) and in the pcre_printint() function (omitted "default"
1595 switch label when the default is to do nothing).
1596
1597 15. Added some code to make it possible, when PCRE is compiled as a C++
1598 library, to replace subject pointers for pcre_exec() with a smart pointer
1599 class, thus making it possible to process discontinuous strings.
1600
1601 16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
1602 much the same function. They were added by different people who were trying
1603 to make PCRE easy to compile on non-Unix systems. It has been suggested
1604 that PCRE_EXPORT be abolished now that there is more automatic apparatus
1605 for compiling on Windows systems. I have therefore replaced it with
1606 PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
1607 defaults to "extern" for C or "extern C" for C++, which works fine on
1608 Unix-like systems. It is now possible to override the value of PCRE_DATA_
1609 SCOPE with something explicit in config.h. In addition:
1610
1611 (a) pcreposix.h still had just "extern" instead of either of these macros;
1612 I have replaced it with PCRE_DATA_SCOPE.
1613
1614 (b) Functions such as _pcre_xclass(), which are internal to the library,
1615 but external in the C sense, all had PCRE_EXPORT in their definitions.
1616 This is apparently wrong for the Windows case, so I have removed it.
1617 (It makes no difference on Unix-like systems.)
1618
1619 17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
1620 of recursive calls to match(). This is different to MATCH_LIMIT because
1621 that limits the total number of calls to match(), not all of which increase
1622 the depth of recursion. Limiting the recursion depth limits the amount of
1623 stack (or heap if NO_RECURSE is set) that is used. The default can be set
1624 when PCRE is compiled, and changed at run time. A patch from Google adds
1625 this functionality to the C++ interface.
1626
1627 18. Changes to the handling of Unicode character properties:
1628
1629 (a) Updated the table to Unicode 4.1.0.
1630
1631 (b) Recognize characters that are not in the table as "Cn" (undefined).
1632
1633 (c) I revised the way the table is implemented to a much improved format
1634 which includes recognition of ranges. It now supports the ranges that
1635 are defined in UnicodeData.txt, and it also amalgamates other
1636 characters into ranges. This has reduced the number of entries in the
1637 table from around 16,000 to around 3,000, thus reducing its size
1638 considerably. I realized I did not need to use a tree structure after
1639 all - a binary chop search is just as efficient. Having reduced the
1640 number of entries, I extended their size from 6 bytes to 8 bytes to
1641 allow for more data.
1642
1643 (d) Added support for Unicode script names via properties such as \p{Han}.
1644
1645 19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
1646 matching that character.
1647
1648 20. When matching a repeated Unicode property with a minimum greater than zero,
1649 (for example \pL{2,}), PCRE could look past the end of the subject if it
1650 reached it while seeking the minimum number of characters. This could
1651 happen only if some of the characters were more than one byte long, because
1652 there is a check for at least the minimum number of bytes.
1653
1654 21. Refactored the implementation of \p and \P so as to be more general, to
1655 allow for more different types of property in future. This has changed the
1656 compiled form incompatibly. Anybody with saved compiled patterns that use
1657 \p or \P will have to recompile them.
1658
1659 22. Added "Any" and "L&" to the supported property types.
1660
1661 23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
1662 but give a compile time error if the value is greater than 0xff.
1663
1664 24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
1665 accidentally not being installed or uninstalled.
1666
1667 25. The pcre.h file was built from pcre.h.in, but the only changes that were
1668 made were to insert the current release number. This seemed silly, because
1669 it made things harder for people building PCRE on systems that don't run
1670 "configure". I have turned pcre.h into a distributed file, no longer built
1671 by "configure", with the version identification directly included. There is
1672 no longer a pcre.h.in file.
1673
1674 However, this change necessitated a change to the pcre-config script as
1675 well. It is built from pcre-config.in, and one of the substitutions was the
1676 release number. I have updated configure.ac so that ./configure now finds
1677 the release number by grepping pcre.h.
1678
1679 26. Added the ability to run the tests under valgrind.
1680
1681
1682 Version 6.4 05-Sep-05
1683 ---------------------
1684
1685 1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
1686 "--" to be printed when multiple files were scanned, even when none of the
1687 -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
1688 consider it to be a bug, and have restored the previous behaviour.
1689
1690 2. A couple of code tidies to get rid of compiler warnings.
1691
1692 3. The pcretest program used to cheat by referring to symbols in the library
1693 whose names begin with _pcre_. These are internal symbols that are not
1694 really supposed to be visible externally, and in some environments it is
1695 possible to suppress them. The cheating is now confined to including
1696 certain files from the library's source, which is a bit cleaner.
1697
1698 4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
1699 file's purpose clearer.
1700
1701 5. Reorganized pcre_ucp_findchar().
1702
1703
1704 Version 6.3 15-Aug-05
1705 ---------------------
1706
1707 1. The file libpcre.pc.in did not have general read permission in the tarball.
1708
1709 2. There were some problems when building without C++ support:
1710
1711 (a) If C++ support was not built, "make install" and "make test" still
1712 tried to test it.
1713
1714 (b) There were problems when the value of CXX was explicitly set. Some
1715 changes have been made to try to fix these, and ...
1716
1717 (c) --disable-cpp can now be used to explicitly disable C++ support.
1718
1719 (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
1720 backslash in a target when C++ was disabled. This confuses some
1721 versions of "make", apparently. Using an intermediate variable solves
1722 this. (Same for CPP_LOBJ.)
1723
1724 3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
1725 (non-Windows) now includes $(CFLAGS) because these flags are sometimes
1726 necessary on certain architectures.
1727
1728 4. Added a setting of -export-symbols-regex to the link command to remove
1729 those symbols that are exported in the C sense, but actually are local
1730 within the library, and not documented. Their names all begin with
1731 "_pcre_". This is not a perfect job, because (a) we have to except some
1732 symbols that pcretest ("illegally") uses, and (b) the facility isn't always
1733 available (and never for static libraries). I have made a note to try to
1734 find a way round (a) in the future.
1735
1736
1737 Version 6.2 01-Aug-05
1738 ---------------------
1739
1740 1. There was no test for integer overflow of quantifier values. A construction
1741 such as {1111111111111111} would give undefined results. What is worse, if
1742 a minimum quantifier for a parenthesized subpattern overflowed and became
1743 negative, the calculation of the memory size went wrong. This could have
1744 led to memory overwriting.
1745
1746 2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
1747
1748 3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
1749 operating environments where this matters.
1750
1751 4. Applied Giuseppe Maxia's patch to add additional features for controlling
1752 PCRE options from within the C++ wrapper.
1753
1754 5. Named capturing subpatterns were not being correctly counted when a pattern
1755 was compiled. This caused two problems: (a) If there were more than 100
1756 such subpatterns, the calculation of the memory needed for the whole
1757 compiled pattern went wrong, leading to an overflow error. (b) Numerical
1758 back references of the form \12, where the number was greater than 9, were
1759 not recognized as back references, even though there were sufficient
1760 previous subpatterns.
1761
1762 6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
1763 versions of gcc, e.g. 2.95.4.
1764
1765
1766 Version 6.1 21-Jun-05
1767 ---------------------
1768
1769 1. There was one reference to the variable "posix" in pcretest.c that was not
1770 surrounded by "#if !defined NOPOSIX".
1771
1772 2. Make it possible to compile pcretest without DFA support, UTF8 support, or
1773 the cross-check on the old pcre_info() function, for the benefit of the
1774 cut-down version of PCRE that is currently imported into Exim.
1775
1776 3. A (silly) pattern starting with (?i)(?-i) caused an internal space
1777 allocation error. I've done the easy fix, which wastes 2 bytes for sensible
1778 patterns that start (?i) but I don't think that matters. The use of (?i) is
1779 just an example; this all applies to the other options as well.
1780
1781 4. Since libtool seems to echo the compile commands it is issuing, the output
1782 from "make" can be reduced a bit by putting "@" in front of each libtool
1783 compile command.
1784
1785 5. Patch from the folks at Google for configure.in to be a bit more thorough
1786 in checking for a suitable C++ installation before trying to compile the
1787 C++ stuff. This should fix a reported problem when a compiler was present,
1788 but no suitable headers.
1789
1790 6. The man pages all had just "PCRE" as their title. I have changed them to
1791 be the relevant file name. I have also arranged that these names are
1792 retained in the file doc/pcre.txt, which is a concatenation in text format
1793 of all the man pages except the little individual ones for each function.
1794
1795 7. The NON-UNIX-USE file had not been updated for the different set of source
1796 files that come with release 6. I also added a few comments about the C++
1797 wrapper.
1798
1799
1800 Version 6.0 07-Jun-05
1801 ---------------------
1802
1803 1. Some minor internal re-organization to help with my DFA experiments.
1804
1805 2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
1806 didn't matter for the library itself when fully configured, but did matter
1807 when compiling without UCP support, or within Exim, where the ucp files are
1808 not imported.
1809
1810 3. Refactoring of the library code to split up the various functions into
1811 different source modules. The addition of the new DFA matching code (see
1812 below) to a single monolithic source would have made it really too
1813 unwieldy, quite apart from causing all the code to be include in a
1814 statically linked application, when only some functions are used. This is
1815 relevant even without the DFA addition now that patterns can be compiled in
1816 one application and matched in another.
1817
1818 The downside of splitting up is that there have to be some external
1819 functions and data tables that are used internally in different modules of
1820 the library but which are not part of the API. These have all had their
1821 names changed to start with "_pcre_" so that they are unlikely to clash
1822 with other external names.
1823
1824 4. Added an alternate matching function, pcre_dfa_exec(), which matches using
1825 a different (DFA) algorithm. Although it is slower than the original
1826 function, it does have some advantages for certain types of matching
1827 problem.
1828
1829 5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
1830 including restarting after a partial match.
1831
1832 6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
1833 defined when compiling for Windows was sent to me. I have put it into the
1834 code, though I have no means of testing or verifying it.
1835
1836 7. Added the pcre_refcount() auxiliary function.
1837
1838 8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
1839 match before or at the first newline in the subject string. In pcretest,
1840 the /f option on a pattern can be used to set this.
1841
1842 9. A repeated \w when used in UTF-8 mode with characters greater than 256
1843 would behave wrongly. This has been present in PCRE since release 4.0.
1844
1845 10. A number of changes to the pcregrep command:
1846
1847 (a) Refactored how -x works; insert ^(...)$ instead of setting
1848 PCRE_ANCHORED and checking the length, in preparation for adding
1849 something similar for -w.
1850
1851 (b) Added the -w (match as a word) option.
1852
1853 (c) Refactored the way lines are read and buffered so as to have more
1854 than one at a time available.
1855
1856 (d) Implemented a pcregrep test script.
1857
1858 (e) Added the -M (multiline match) option. This allows patterns to match
1859 over several lines of the subject. The buffering ensures that at least
1860 8K, or the rest of the document (whichever is the shorter) is available
1861 for matching (and similarly the previous 8K for lookbehind assertions).
1862
1863 (f) Changed the --help output so that it now says
1864
1865 -w, --word-regex(p)
1866
1867 instead of two lines, one with "regex" and the other with "regexp"
1868 because that confused at least one person since the short forms are the
1869 same. (This required a bit of code, as the output is generated
1870 automatically from a table. It wasn't just a text change.)
1871
1872 (g) -- can be used to terminate pcregrep options if the next thing isn't an
1873 option but starts with a hyphen. Could be a pattern or a path name
1874 starting with a hyphen, for instance.
1875
1876 (h) "-" can be given as a file name to represent stdin.
1877
1878 (i) When file names are being printed, "(standard input)" is used for
1879 the standard input, for compatibility with GNU grep. Previously
1880 "<stdin>" was used.
1881
1882 (j) The option --label=xxx can be used to supply a name to be used for
1883 stdin when file names are being printed. There is no short form.
1884
1885 (k) Re-factored the options decoding logic because we are going to add
1886 two more options that take data. Such options can now be given in four
1887 different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
1888
1889 (l) Added the -A, -B, and -C options for requesting that lines of context
1890 around matches be printed.
1891
1892 (m) Added the -L option to print the names of files that do not contain
1893 any matching lines, that is, the complement of -l.
1894
1895 (n) The return code is 2 if any file cannot be opened, but pcregrep does
1896 continue to scan other files.
1897
1898 (o) The -s option was incorrectly implemented. For compatibility with other
1899 greps, it now suppresses the error message for a non-existent or non-
1900 accessible file (but not the return code). There is a new option called
1901 -q that suppresses the output of matching lines, which was what -s was
1902 previously doing.
1903
1904 (p) Added --include and --exclude options to specify files for inclusion
1905 and exclusion when recursing.
1906
1907 11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
1908 Hopefully, it now does.
1909
1910 12. Missing cast in pcre_study().
1911
1912 13. Added an "uninstall" target to the makefile.
1913
1914 14. Replaced "extern" in the function prototypes in Makefile.in with
1915 "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
1916 world, but is set differently for Windows.
1917
1918 15. Added a second compiling function called pcre_compile2(). The only
1919 difference is that it has an extra argument, which is a pointer to an
1920 integer error code. When there is a compile-time failure, this is set
1921 non-zero, in addition to the error test pointer being set to point to an
1922 error message. The new argument may be NULL if no error number is required
1923 (but then you may as well call pcre_compile(), which is now just a
1924 wrapper). This facility is provided because some applications need a
1925 numeric error indication, but it has also enabled me to tidy up the way
1926 compile-time errors are handled in the POSIX wrapper.
1927
1928 16. Added VPATH=.libs to the makefile; this should help when building with one
1929 prefix path and installing with another. (Or so I'm told by someone who
1930 knows more about this stuff than I do.)
1931
1932 17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
1933 passes PCRE_DOTALL to the pcre_compile() function, making the "." character
1934 match everything, including newlines. This is not POSIX-compatible, but
1935 somebody wanted the feature. From pcretest it can be activated by using
1936 both the P and the s flags.
1937
1938 18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
1939
1940 19. libpcre.pc was being incorrectly installed as executable.
1941
1942 20. A couple of places in pcretest check for end-of-line by looking for '\n';
1943 it now also looks for '\r' so that it will work unmodified on Windows.
1944
1945 21. Added Google's contributed C++ wrapper to the distribution.
1946
1947 22. Added some untidy missing memory free() calls in pcretest, to keep
1948 Electric Fence happy when testing.
1949
1950
1951
1952 Version 5.0 13-Sep-04
1953 ---------------------
1954
1955 1. Internal change: literal characters are no longer packed up into items
1956 containing multiple characters in a single byte-string. Each character
1957 is now matched using a separate opcode. However, there may be more than one
1958 byte in the character in UTF-8 mode.
1959
1960 2. The pcre_callout_block structure has two new fields: pattern_position and
1961 next_item_length. These contain the offset in the pattern to the next match
1962 item, and its length, respectively.
1963
1964 3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
1965 insertion of callouts before each pattern item. Added the /C option to
1966 pcretest to make use of this.
1967
1968 4. On the advice of a Windows user, the lines
1969
1970 #if defined(_WIN32) || defined(WIN32)
1971 _setmode( _fileno( stdout ), 0x8000 );
1972 #endif /* defined(_WIN32) || defined(WIN32) */
1973
1974 have been added to the source of pcretest. This apparently does useful
1975 magic in relation to line terminators.
1976
1977 5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
1978 for the benefit of those environments where the "b" makes a difference.
1979
1980 6. The icc compiler has the same options as gcc, but "configure" doesn't seem
1981 to know about it. I have put a hack into configure.in that adds in code
1982 to set GCC=yes if CC=icc. This seems to end up at a point in the
1983 generated configure script that is early enough to affect the setting of
1984 compiler options, which is what is needed, but I have no means of testing
1985 whether it really works. (The user who reported this had patched the
1986 generated configure script, which of course I cannot do.)
1987
1988 LATER: After change 22 below (new libtool files), the configure script
1989 seems to know about icc (and also ecc). Therefore, I have commented out
1990 this hack in configure.in.
1991
1992 7. Added support for pkg-config (2 patches were sent in).
1993
1994 8. Negated POSIX character classes that used a combination of internal tables
1995 were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
1996 [[:^ascii]]. Typically, they would match almost any characters. The other
1997 POSIX classes were not broken in this way.
1998
1999 9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
2000 to find the match, as PCRE was deluded into thinking that the match had to
2001 start at the start point or following a newline. The same bug applied to
2002 patterns with negative forward assertions or any backward assertions
2003 preceding ".*" at the start, unless the pattern required a fixed first
2004 character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
2005
2006 10. In UTF-8 mode, when moving forwards in the subject after a failed match
2007 starting at the last subject character, bytes beyond the end of the subject
2008 string were read.
2009
2010 11. Renamed the variable "class" as "classbits" to make life easier for C++
2011 users. (Previously there was a macro definition, but it apparently wasn't
2012 enough.)
2013
2014 12. Added the new field "tables" to the extra data so that tables can be passed
2015 in at exec time, or the internal tables can be re-selected. This allows
2016 a compiled regex to be saved and re-used at a later time by a different
2017 program that might have everything at different addresses.
2018
2019 13. Modified the pcre-config script so that, when run on Solaris, it shows a
2020 -R library as well as a -L library.
2021
2022 14. The debugging options of pcretest (-d on the command line or D on a
2023 pattern) showed incorrect output for anything following an extended class
2024 that contained multibyte characters and which was followed by a quantifier.
2025
2026 15. Added optional support for general category Unicode character properties
2027 via the \p, \P, and \X escapes. Unicode property support implies UTF-8
2028 support. It adds about 90K to the size of the library. The meanings of the
2029 inbuilt class escapes such as \d and \s have NOT been changed.
2030
2031 16. Updated pcredemo.c to include calls to free() to release the memory for the
2032 compiled pattern.
2033
2034 17. The generated file chartables.c was being created in the source directory
2035 instead of in the building directory. This caused the build to fail if the
2036 source directory was different from the building directory, and was
2037 read-only.
2038
2039 18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
2040 file. No doubt somebody will tell me if they don't make sense... Also added
2041 Dan Mooney's comments about building on OpenVMS.
2042
2043 19. Added support for partial matching via the PCRE_PARTIAL option for
2044 pcre_exec() and the \P data escape in pcretest.
2045
2046 20. Extended pcretest with 3 new pattern features:
2047
2048 (i) A pattern option of the form ">rest-of-line" causes pcretest to
2049 write the compiled pattern to the file whose name is "rest-of-line".
2050 This is a straight binary dump of the data, with the saved pointer to
2051 the character tables forced to be NULL. The study data, if any, is
2052 written too. After writing, pcretest reads a new pattern.
2053
2054 (ii) If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
2055 compiled pattern from the given file. There must not be any
2056 occurrences of "<" in the file name (pretty unlikely); if there are,
2057 pcretest will instead treat the initial "<" as a pattern delimiter.
2058 After reading in the pattern, pcretest goes on to read data lines as
2059 usual.
2060
2061 (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
2062 and 16-bit fields in a compiled pattern, to simulate a pattern that
2063 was compiled on a host of opposite endianness.
2064
2065 21. The pcre-exec() function can now cope with patterns that were compiled on
2066 hosts of opposite endianness, with this restriction:
2067
2068 As for any compiled expression that is saved and used later, the tables
2069 pointer field cannot be preserved; the extra_data field in the arguments
2070 to pcre_exec() should be used to pass in a tables address if a value
2071 other than the default internal tables were used at compile time.
2072
2073 22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
2074 now diagnosed as an error. Previously, most of the time, a negative number
2075 would have been treated as zero, but if in addition "ovector" was passed as
2076 NULL, a crash could occur.
2077
2078 23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
2079 new versions from the libtool 1.5 distribution (the last one is a copy of
2080 a file called libtool.m4). This seems to have fixed the need to patch
2081 "configure" to support Darwin 1.3 (which I used to do). However, I still
2082 had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
2083 workstation).
2084
2085 24. Changed the PCRE licence to be the more standard "BSD" licence.
2086
2087
2088 Version 4.5 01-Dec-03
2089 ---------------------
2090
2091 1. There has been some re-arrangement of the code for the match() function so
2092 that it can be compiled in a version that does not call itself recursively.
2093 Instead, it keeps those local variables that need separate instances for
2094 each "recursion" in a frame on the heap, and gets/frees frames whenever it
2095 needs to "recurse". Keeping track of where control must go is done by means
2096 of setjmp/longjmp. The whole thing is implemented by a set of macros that
2097 hide most of the details from the main code, and operates only if
2098 NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
2099 "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
2100 operating.
2101
2102 To make it easier for callers to provide specially tailored get/free
2103 functions for this usage, two new functions, pcre_stack_malloc, and
2104 pcre_stack_free, are used. They are always called in strict stacking order,
2105 and the size of block requested is always the same.
2106
2107 The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
2108 PCRE has been compiled to use the stack or the heap for recursion. The
2109 -C option of pcretest uses this to show which version is compiled.
2110
2111 A new data escape \S, is added to pcretest; it causes the amounts of store
2112 obtained and freed by both kinds of malloc/free at match time to be added
2113 to the output.
2114
2115 2. Changed the locale test to use "fr_FR" instead of "fr" because that's
2116 what's available on my current Linux desktop machine.
2117
2118 3. When matching a UTF-8 string, the test for a valid string at the start has
2119 been extended. If start_offset is not zero, PCRE now checks that it points
2120 to a byte that is the start of a UTF-8 character. If not, it returns
2121 PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
2122 this is necessary because there may be backward assertions in the pattern.
2123 When matching the same subject several times, it may save resources to use
2124 PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
2125
2126 4. The code for checking the validity of UTF-8 strings has been tightened so
2127 that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
2128 containing "overlong sequences".
2129
2130 5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
2131 I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
2132 should have been "&", but it just so happened that all the cases this let
2133 through by mistake were picked up later in the function.
2134
2135 6. I had used a variable called "isblank" - this is a C99 function, causing
2136 some compilers to warn. To avoid this, I renamed it (as "blankclass").
2137
2138 7. Cosmetic: (a) only output another newline at the end of pcretest if it is
2139 prompting; (b) run "./pcretest /dev/null" at the start of the test script
2140 so the version is shown; (c) stop "make test" echoing "./RunTest".
2141
2142 8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
2143
2144 9. The prototype for memmove() for systems that don't have it was using
2145 size_t, but the inclusion of the header that defines size_t was later. I've
2146 moved the #includes for the C headers earlier to avoid this.
2147
2148 10. Added some adjustments to the code to make it easier to compiler on certain
2149 special systems:
2150
2151 (a) Some "const" qualifiers were missing.
2152 (b) Added the macro EXPORT before all exported functions; by default this
2153 is defined to be empty.
2154 (c) Changed the dftables auxiliary program (that builds chartables.c) so
2155 that it reads its output file name as an argument instead of writing
2156 to the standard output and assuming this can be redirected.
2157
2158 11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
2159 class containing characters with values greater than 255, PCRE compilation
2160 went into a loop.
2161
2162 12. A recursive reference to a subpattern that was within another subpattern
2163 that had a minimum quantifier of zero caused PCRE to crash. For example,
2164 (x(y(?2))z)? provoked this bug with a subject that got as far as the
2165 recursion. If the recursively-called subpattern itself had a zero repeat,
2166 that was OK.
2167
2168 13. In pcretest, the buffer for reading a data line was set at 30K, but the
2169 buffer into which it was copied (for escape processing) was still set at
2170 1024, so long lines caused crashes.
2171
2172 14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
2173 "internal error: code overflow...". This applied to any character class
2174 that was followed by a possessive quantifier.
2175
2176 15. Modified the Makefile to add libpcre.la as a prerequisite for
2177 libpcreposix.la because I was told this is needed for a parallel build to
2178 work.
2179
2180 16. If a pattern that contained .* following optional items at the start was
2181 studied, the wrong optimizing data was generated, leading to matching
2182 errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
2183 matching string must start with a or b or c. The correct conclusion for
2184 this pattern is that a match can start with any character.
2185
2186
2187 Version 4.4 13-Aug-03
2188 ---------------------
2189
2190 1. In UTF-8 mode, a character class containing characters with values between
2191 127 and 255 was not handled correctly if the compiled pattern was studied.
2192 In fixing this, I have also improved the studying algorithm for such
2193 classes (slightly).
2194
2195 2. Three internal functions had redundant arguments passed to them. Removal
2196 might give a very teeny performance improvement.
2197
2198 3. Documentation bug: the value of the capture_top field in a callout is *one
2199 more than* the number of the hightest numbered captured substring.
2200
2201 4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
2202 in incorrectly linking with a previously installed version. They now link
2203 explicitly with libpcre.la.
2204
2205 5. configure.in no longer needs to recognize Cygwin specially.
2206
2207 6. A problem in pcre.in for Windows platforms is fixed.
2208
2209 7. If a pattern was successfully studied, and the -d (or /D) flag was given to
2210 pcretest, it used to include the size of the study block as part of its
2211 output. Unfortunately, the structure contains a field that has a different
2212 size on different hardware architectures. This meant that the tests that
2213 showed this size failed. As the block is currently always of a fixed size,
2214 this information isn't actually particularly useful in pcretest output, so
2215 I have just removed it.
2216
2217 8. Three pre-processor statements accidentally did not start in column 1.
2218 Sadly, there are *still* compilers around that complain, even though
2219 standard C has not required this for well over a decade. Sigh.
2220
2221 9. In pcretest, the code for checking callouts passed small integers in the
2222 callout_data field, which is a void * field. However, some picky compilers
2223 complained about the casts involved for this on 64-bit systems. Now
2224 pcretest passes the address of the small integer instead, which should get
2225 rid of the warnings.
2226
2227 10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
2228 both compile and run time, and gives an error if an invalid UTF-8 sequence
2229 is found. There is a option for disabling this check in cases where the
2230 string is known to be correct and/or the maximum performance is wanted.
2231
2232 11. In response to a bug report, I changed one line in Makefile.in from
2233
2234 -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
2235 to
2236 -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
2237
2238 to look similar to other lines, but I have no way of telling whether this
2239 is the right thing to do, as I do not use Windows. No doubt I'll get told
2240 if it's wrong...
2241
2242
2243 Version 4.3 21-May-03
2244 ---------------------
2245
2246 1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
2247 Makefile.
2248
2249 2. Some refactoring to improve the quality of the code:
2250
2251 (i) The utf8_table... variables are now declared "const".
2252
2253 (ii) The code for \cx, which used the "case flipping" table to upper case
2254 lower case letters, now just substracts 32. This is ASCII-specific,
2255 but the whole concept of \cx is ASCII-specific, so it seems
2256 reasonable.
2257
2258 (iii) PCRE was using its character types table to recognize decimal and
2259 hexadecimal digits in the pattern. This is silly, because it handles
2260 only 0-9, a-f, and A-F, but the character types table is locale-
2261 specific, which means strange things might happen. A private
2262 table is now used for this - though it costs 256 bytes, a table is
2263 much faster than multiple explicit tests. Of course, the standard
2264 character types table is still used for matching digits in subject
2265 strings against \d.
2266
2267 (iv) Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
2268 ending in _t are). So I've renamed it as ESC_tee.
2269
2270 3. The first argument for regexec() in the POSIX wrapper should have been
2271 defined as "const".
2272
2273 4. Changed pcretest to use malloc() for its buffers so that they can be
2274 Electric Fenced for debugging.
2275
2276 5. There were several places in the code where, in UTF-8 mode, PCRE would try
2277 to read one or more bytes before the start of the subject string. Often this
2278 had no effect on PCRE's behaviour, but in some circumstances it could
2279 provoke a segmentation fault.
2280
2281 6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
2282 to try to read one or more bytes before the start of the subject string.
2283
2284 7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
2285 UTF-8 support could misbehave in various ways if the subject string
2286 contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
2287 area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
2288 back over UTF-8 characters.)
2289
2290
2291 Version 4.2 14-Apr-03
2292 ---------------------
2293
2294 1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
2295
2296 2. Changes to the building process, supplied by Ronald Landheer-Cieslak
2297 [ON_WINDOWS]: new variable, "#" on non-Windows platforms
2298 [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
2299 [WIN_PREFIX]: new variable, "cyg" for Cygwin
2300 * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
2301 and BUILD_EXEEXT
2302 Note: automatic setting of the BUILD variables is not yet working
2303 set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
2304 compile-time but not at link-time
2305 [LINK]: use for linking executables only
2306 make different versions for Windows and non-Windows
2307 [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
2308 libraries
2309 [LINK_FOR_BUILD]: new variable
2310 [OBJEXT]: use throughout
2311 [EXEEXT]: use throughout
2312 <winshared>: new target
2313 <wininstall>: new target
2314 <dftables.o>: use native compiler
2315 <dftables>: use native linker
2316 <install>: handle Windows platform correctly
2317 <clean>: ditto
2318 <check>: ditto
2319 copy DLL to top builddir before testing
2320
2321 As part of these changes, -no-undefined was removed again. This was reported
2322 to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
2323 in any case.
2324
2325 3. Some tidies to get rid of compiler warnings:
2326
2327 . In the match_data structure, match_limit was an unsigned long int, whereas
2328 match_call_count was an int. I've made them both unsigned long ints.
2329
2330 . In pcretest the fact that a const uschar * doesn't automatically cast to
2331 a void * provoked a warning.
2332
2333 . Turning on some more compiler warnings threw up some "shadow" variables
2334 and a few more missing casts.
2335
2336 4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
2337 option, a class that contained a single character with a value between 128
2338 and 255 (e.g. /[\xFF]/) caused PCRE to crash.
2339
2340 5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
2341 option, a class that contained several characters, but with at least one
2342 whose value was between 128 and 255 caused PCRE to crash.
2343
2344
2345 Version 4.1 12-Mar-03
2346 ---------------------
2347
2348 1. Compiling with gcc -pedantic found a couple of places where casts were
2349 needed, and a string in dftables.c that was longer than standard compilers are
2350 required to support.
2351
2352 2. Compiling with Sun's compiler found a few more places where the code could
2353 be tidied up in order to avoid warnings.
2354
2355 3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
2356 first of these names is deprecated in the latest Autoconf in favour of the name
2357 CC_FOR_BUILD, because "host" is typically used to mean the system on which the
2358 compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
2359 analogy I have changed it to CFLAGS_FOR_BUILD.
2360
2361 4. Added -no-undefined to the linking command in the Makefile, because this is
2362 apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
2363 linking step for the pcreposix library.
2364
2365 5. PCRE was failing to diagnose the case of two named groups with the same
2366 name.
2367
2368 6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
2369 literal character that is needed in the subject for a match, and scans along to
2370 ensure that it is present before embarking on the full matching process. This
2371 saves time in cases of nested unlimited repeats that are never going to match.
2372 Problem: the scan can take a lot of time if the subject is very long (e.g.
2373 megabytes), thus penalizing straightforward matches. It is now done only if the
2374 amount of subject to be scanned is less than 1000 bytes.
2375
2376 7. A lesser problem with the same optimization is that it was recording the
2377 first character of an anchored pattern as "needed", thus provoking a search
2378 right along the subject, even when the first match of the pattern was going to
2379 fail. The "needed" character is now not set for anchored patterns, unless it
2380 follows something in the pattern that is of non-fixed length. Thus, it still
2381 fulfils its original purpose of finding quick non-matches in cases of nested
2382 unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
2383
2384
2385 Version 4.0 17-Feb-03
2386 ---------------------
2387
2388 1. If a comment in an extended regex that started immediately after a meta-item
2389 extended to the end of string, PCRE compiled incorrect data. This could lead to
2390 all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
2391
2392 2. Moved to autoconf 2.53 and libtool 1.4.2.
2393
2394 3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
2395 the special perltest8 script is no longer needed - all the tests can be run
2396 from a single perltest script.
2397
2398 4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
2399 by \s. It has now been removed in PCRE. This means it isn't recognized as
2400 whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
2401 class [:space:] *does* include VT, thereby creating a mess.
2402
2403 5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
2404 space and tab.
2405
2406 6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
2407 its new features into the main test script, reducing the number of scripts.
2408
2409 7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
2410 were backward compatible, and made the (?i) apply to the whole pattern, as if
2411 /i were given. Now it behaves more logically, and applies the option setting
2412 only to what follows. PCRE has been changed to follow suit. However, if it
2413 finds options settings right at the start of the pattern, it extracts them into
2414 the global options, as before. Thus, they show up in the info data.
2415
2416 8. Added support for the \Q...\E escape sequence. Characters in between are
2417 treated as literals. This is slightly different from Perl in that $ and @ are
2418 also handled as literals inside the quotes. In Perl, they will cause variable
2419 interpolation. Note the following examples:
2420
2421 Pattern PCRE matches Perl matches
2422
2423 \Qabc$xyz\E abc$xyz abc followed by the contents of $xyz
2424 \Qabc\$xyz\E abc\$xyz abc\$xyz
2425 \Qabc\E\$\Qxyz\E abc$xyz abc$xyz
2426
2427 For compatibility with Perl, \Q...\E sequences are recognized inside character
2428 classes as well as outside them.
2429
2430 9. Re-organized 3 code statements in pcretest to avoid "overflow in
2431 floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
2432 (size_t) cast to one statement in pcretest and one in pcreposix to avoid
2433 signed/unsigned warnings.
2434
2435 10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
2436 option for pcretest, so I've replaced it by a simple function that does just
2437 that job.
2438
2439 11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
2440 "pcregrep -".
2441
2442 12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
2443 Java package. This provides some syntactic sugar for simple cases of what my
2444 documentation calls "once-only subpatterns". A pattern such as x*+ is the same
2445 as (?>x*). In other words, if what is inside (?>...) is just a single repeated
2446 item, you can use this simplified notation. Note that only makes sense with
2447 greedy quantifiers. Consequently, the use of the possessive quantifier forces
2448 greediness, whatever the setting of the PCRE_UNGREEDY option.
2449
2450 13. A change of greediness default within a pattern was not taking effect at
2451 the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
2452 subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
2453 was abstracted outside.
2454
2455 14. PCRE now supports the \G assertion. It is true when the current matching
2456 position is at the start point of the match. This differs from \A when the
2457 starting offset is non-zero. Used with the /g option of pcretest (or similar
2458 code), it works in the same way as it does for Perl's /g option. If all
2459 alternatives of a regex begin with \G, the expression is anchored to the start
2460 match position, and the "anchored" flag is set in the compiled expression.
2461
2462 15. Some bugs concerning the handling of certain option changes within patterns
2463 have been fixed. These applied to options other than (?ims). For example,
2464 "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
2465 been the other way round. Some of this was related to change 7 above.
2466
2467 16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
2468 features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
2469 and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
2470 POSIX classes only within a class (e.g. /[[:alpha:]]/).
2471
2472 17. Added support for Perl's \C escape. This matches one byte, even in UTF8
2473 mode. Unlike ".", it always matches newline, whatever the setting of
2474 PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
2475 assertions. Perl allows it, but it doesn't (in general) work because it can't
2476 calculate the length of the lookbehind. At least, that's the case for Perl
2477 5.8.0 - I've been told they are going to document that it doesn't work in
2478 future.
2479
2480 18. Added an error diagnosis for escapes that PCRE does not support: these are
2481 \L, \l, \N, \P, \p, \U, \u, and \X.
2482
2483 19. Although correctly diagnosing a missing ']' in a character class, PCRE was
2484 reading past the end of the pattern in cases such as /[abcd/.
2485
2486 20. PCRE was getting more memory than necessary for patterns with classes that
2487 contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
2488
2489 21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
2490 compiling PCRE for use with Virtual Pascal.
2491
2492 22. Small fix to the Makefile to make it work properly if the build is done
2493 outside the source tree.
2494
2495 23. Added a new extension: a condition to go with recursion. If a conditional
2496 subpattern starts with (?(R) the "true" branch is used if recursion has
2497 happened, whereas the "false" branch is used only at the top level.
2498
2499 24. When there was a very long string of literal characters (over 255 bytes
2500 without UTF support, over 250 bytes with UTF support), the computation of how
2501 much memory was required could be incorrect, leading to segfaults or other
2502 strange effects.
2503
2504 25. PCRE was incorrectly assuming anchoring (either to start of subject or to
2505 start of line for a non-DOTALL pattern) when a pattern started with (.*) and
2506 there was a subsequent back reference to those brackets. This meant that, for
2507 example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
2508 possible to check for precisely this case. All we can do is abandon the
2509 optimization if .* occurs inside capturing brackets when there are any back
2510 references whatsoever. (See below for a better fix that came later.)
2511
2512 26. The handling of the optimization for finding the first character of a
2513 non-anchored pattern, and for finding a character that is required later in the
2514 match were failing in some cases. This didn't break the matching; it just
2515 failed to optimize when it could. The way this is done has been re-implemented.
2516
2517 27. Fixed typo in error message for invalid (?R item (it said "(?p").
2518
2519 28. Added a new feature that provides some of the functionality that Perl
2520 provides with (?{...}). The facility is termed a "callout". The way it is done
2521 in PCRE is for the caller to provide an optional function, by setting
2522 pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
2523 global variable. By default it is unset, which disables all calling out. To get
2524 the function called, the regex must include (?C) at appropriate points. This
2525 is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
2526 This provides a means of identifying different callout points. When PCRE
2527 reaches such a point in the regex, if pcre_callout has been set, the external
2528 function is called. It is provided with data in a structure called
2529 pcre_callout_block, which is defined in pcre.h. If the function returns 0,
2530 matching continues; if it returns a non-zero value, the match at the current
2531 point fails. However, backtracking will occur if possible. [This was changed
2532 later and other features added - see item 49 below.]
2533
2534 29. pcretest is upgraded to test the callout functionality. It provides a
2535 callout function that displays information. By default, it shows the start of
2536 the match and the current position in the text. There are some new data escapes
2537 to vary what happens:
2538
2539 \C+ in addition, show current contents of captured substrings
2540 \C- do not supply a callout function
2541 \C!n return 1 when callout number n is reached
2542 \C!n!m return 1 when callout number n is reached for the mth time
2543
2544 30. If pcregrep was called with the -l option and just a single file name, it
2545 output "<stdin>" if a match was found, instead of the file name.
2546
2547 31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
2548 slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
2549 pcre_exec(). This saves a malloc/free per call. The default value of
2550 POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
2551 when configuring.
2552
2553 32. The default maximum size of a compiled pattern is 64K. There have been a
2554 few cases of people hitting this limit. The code now uses macros to handle the
2555 storing of links as offsets within the compiled pattern. It defaults to 2-byte
2556 links, but this can be changed to 3 or 4 bytes by --with-link-size when
2557 configuring. Tests 2 and 5 work only with 2-byte links because they output
2558 debugging information about compiled patterns.
2559
2560 33. Internal code re-arrangements:
2561
2562 (a) Moved the debugging function for printing out a compiled regex into
2563 its own source file (printint.c) and used #include to pull it into
2564 pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
2565 separate copies.
2566
2567 (b) Defined the list of op-code names for debugging as a macro in
2568 internal.h so that it is next to the definition of the opcodes.
2569
2570 (c) Defined a table of op-code lengths for simpler skipping along compiled
2571 code. This is again a macro in internal.h so that it is next to the
2572 definition of the opcodes.
2573
2574 34. Added support for recursive calls to individual subpatterns, along the
2575 lines of Robin Houston's patch (but implemented somewhat differently).
2576
2577 35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
2578 allow it to read and process whole directories in Win32. This code was
2579 contributed by Lionel Fourquaux; it has not been tested by me.
2580
2581 36. Added support for named subpatterns. The Python syntax (?P<name>...) is
2582 used to name a group. Names consist of alphanumerics and underscores, and must
2583 be unique. Back references use the syntax (?P=name) and recursive calls use
2584 (?P>name) which is a PCRE extension to the Python extension. Groups still have
2585 numbers. The function pcre_fullinfo() can be used after compilation to extract
2586 a name/number map. There are three relevant calls:
2587
2588 PCRE_INFO_NAMEENTRYSIZE yields the size of each entry in the map
2589 PCRE_INFO_NAMECOUNT yields the number of entries
2590 PCRE_INFO_NAMETABLE yields a pointer to the map.
2591
2592 The map is a vector of fixed-size entries. The size of each entry depends on
2593 the length of the longest name used. The first two bytes of each entry are the
2594 group number, most significant byte first. There follows the corresponding
2595 name, zero terminated. The names are in alphabetical order.
2596
2597 37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
2598 case instead of 255. Making it the same both with and without UTF-8 support
2599 means that the same test output works with both.
2600
2601 38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
2602 calling malloc() with a zero argument.
2603
2604 39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
2605 optimization. I've improved things by keeping a bitmap of backreferences with
2606 numbers 1-31 so that if .* occurs inside capturing brackets that are not in
2607 fact referenced, the optimization can be applied. It is unlikely that a
2608 relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
2609 the match to follow \n) will appear inside brackets with a number greater than
2610 31, but if it does, any back reference > 31 suppresses the optimization.
2611
2612 40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
2613 of disabling numbered capturing parentheses. Any opening parenthesis that is
2614 not followed by ? behaves as if it were followed by ?: but named parentheses
2615 can still be used for capturing (and they will acquire numbers in the usual
2616 way).
2617
2618 41. Redesigned the return codes from the match() function into yes/no/error so
2619 that errors can be passed back from deep inside the nested calls. A malloc
2620 failure while inside a recursive subpattern call now causes the
2621 PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
2622
2623 42. It is now possible to set a limit on the number of times the match()
2624 function is called in a call to pcre_exec(). This facility makes it possible to
2625 limit the amount of recursion and backtracking, though not in a directly
2626 obvious way, because the match() function is used in a number of different
2627 circumstances. The count starts from zero for each position in the subject
2628 string (for non-anchored patterns). The default limit is, for compatibility, a
2629 large number, namely 10 000 000. You can change this in two ways:
2630
2631 (a) When configuring PCRE before making, you can use --with-match-limit=n
2632 to set a default value for the compiled library.
2633
2634 (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
2635 a different value is set. See 45 below.
2636
2637 If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
2638
2639 43. Added a new function pcre_config(int, void *) to enable run-time extraction
2640 of things that can be changed at compile time. The first argument specifies
2641 what is wanted and the second points to where the information is to be placed.
2642 The current list of available information is:
2643
2644 PCRE_CONFIG_UTF8
2645
2646 The output is an integer that is set to one if UTF-8 support is available;
2647 otherwise it is set to zero.
2648
2649 PCRE_CONFIG_NEWLINE
2650
2651 The output is an integer that it set to the value of the code that is used for
2652 newline. It is either LF (10) or CR (13).
2653
2654 PCRE_CONFIG_LINK_SIZE
2655
2656 The output is an integer that contains the number of bytes used for internal
2657 linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
2658
2659 PCRE_CONFIG_POSIX_MALLOC_THRESHOLD
2660
2661 The output is an integer that contains the threshold above which the POSIX
2662 interface uses malloc() for output vectors. See item 31 above.
2663
2664 PCRE_CONFIG_MATCH_LIMIT
2665
2666 The output is an unsigned integer that contains the default limit of the number
2667 of match() calls in a pcre_exec() execution. See 42 above.
2668
2669 44. pcretest has been upgraded by the addition of the -C option. This causes it
2670 to extract all the available output from the new pcre_config() function, and to
2671 output it. The program then exits immediately.
2672
2673 45. A need has arisen to pass over additional data with calls to pcre_exec() in
2674 order to support additional features. One way would have been to define
2675 pcre_exec2() (for example) with extra arguments, but this would not have been
2676 extensible, and would also have required all calls to the original function to
2677 be mapped to the new one. Instead, I have chosen to extend the mechanism that
2678 is used for passing in "extra" data from pcre_study().
2679
2680 The pcre_extra structure is now exposed and defined in pcre.h. It currently
2681 contains the following fields:
2682
2683 flags a bitmap indicating which of the following fields are set
2684 study_data opaque data from pcre_study()
2685 match_limit a way of specifying a limit on match() calls for a specific
2686 call to pcre_exec()
2687 callout_data data for callouts (see 49 below)
2688
2689 The flag bits are also defined in pcre.h, and are
2690
2691 PCRE_EXTRA_STUDY_DATA
2692 PCRE_EXTRA_MATCH_LIMIT
2693 PCRE_EXTRA_CALLOUT_DATA
2694
2695 The pcre_study() function now returns one of these new pcre_extra blocks, with
2696 the actual study data pointed to by the study_data field, and the
2697 PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
2698 before. That is, this change is entirely upwards-compatible and requires no
2699 change to existing code.
2700
2701 If you want to pass in additional data to pcre_exec(), you can either place it
2702 in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
2703 block.
2704
2705 46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
2706 data string contains the escape sequence \M, pcretest calls pcre_exec() several
2707 times with different match limits, until it finds the minimum value needed for
2708 pcre_exec() to complete. The value is then output. This can be instructive; for
2709 most simple matches the number is quite small, but for pathological cases it
2710 gets very large very quickly.
2711
2712 47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
2713 returns the size of the data block pointed to by the study_data field in a
2714 pcre_extra block, that is, the value that was passed as the argument to
2715 pcre_malloc() when PCRE was getting memory in which to place the information
2716 created by pcre_study(). The fourth argument should point to a size_t variable.
2717 pcretest has been extended so that this information is shown after a successful
2718 pcre_study() call when information about the compiled regex is being displayed.
2719
2720 48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
2721 because what follows is always an absolute path. (Later: it turns out that this
2722 is more than cosmetic for MinGW, because it doesn't like empty path
2723 components.)
2724
2725 49. Some changes have been made to the callout feature (see 28 above):
2726
2727 (i) A callout function now has three choices for what it returns:
2728
2729 0 => success, carry on matching
2730 > 0 => failure at this point, but backtrack if possible
2731 < 0 => serious error, return this value from pcre_exec()
2732
2733 Negative values should normally be chosen from the set of PCRE_ERROR_xxx
2734 values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
2735 "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
2736 use by callout functions. It will never be used by PCRE itself.
2737
2738 (ii) The pcre_extra structure (see 45 above) has a void * field called
2739 callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
2740 pcre_callout_block structure has a field of the same name. The contents of
2741 the field passed in the pcre_extra structure are passed to the callout
2742 function in the corresponding field in the callout block. This makes it
2743 easier to use the same callout-containing regex from multiple threads. For
2744 testing, the pcretest program has a new data escape
2745
2746 \C*n pass the number n (may be negative) as callout_data
2747
2748 If the callout function in pcretest receives a non-zero value as
2749 callout_data, it returns that value.
2750
2751 50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
2752 there were some redundant $(CFLAGS) in commands that are now specified as
2753 $(LINK), which already includes $(CFLAGS).
2754
2755 51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
2756 has been compiled with UTF-8 support *and* pcre_compile() has been compiled
2757 with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
2758 one-byte characters throughout. Note that case-insensitive matching applies
2759 only to characters whose values are less than 256. PCRE doesn't support the
2760 notion of cases for higher-valued characters.
2761
2762 (i) A character class whose characters are all within 0-255 is handled as
2763 a bit map, and the map is inverted for negative classes. Previously, a
2764 character > 255 always failed to match such a class; however it should
2765 match if the class was a negative one (e.g. [^ab]). This has been fixed.
2766
2767 (ii) A negated character class with a single character < 255 is coded as
2768 "not this character" (OP_NOT). This wasn't working properly when the test
2769 character was multibyte, either singly or repeated.
2770
2771 (iii) Repeats of multibyte characters are now handled correctly in UTF-8
2772 mode, for example: \x{100}{2,3}.
2773
2774 (iv) The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
2775 singly or repeated) now correctly test multibyte characters. However,
2776 PCRE doesn't recognize any characters with values greater than 255 as
2777 digits, spaces, or word characters. Such characters always match \D, \S,
2778 and \W, and never match \d, \s, or \w.
2779
2780 (v) Classes may now contain characters and character ranges with values
2781 greater than 255. For example: [ab\x{100}-\x{400}].
2782
2783 (vi) pcregrep now has a --utf-8 option (synonym -u) which makes it call
2784 PCRE in UTF-8 mode.
2785
2786 52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
2787 PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
2788 retained for backwards compatibility. (Note that LASTLITERAL is also a byte
2789 value.)
2790
2791 53. The single man page has become too large. I have therefore split it up into
2792 a number of separate man pages. These also give rise to individual HTML pages;
2793 these are now put in a separate directory, and there is an index.html page that
2794 lists them all. Some hyperlinking between the pages has been installed.
2795
2796 54. Added convenience functions for handling named capturing parentheses.
2797
2798 55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
2799 aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
2800 true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
2801 are faulted.
2802
2803 56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
2804 calling configure. These values are used when compiling the dftables.c program
2805 which is run to generate the source of the default character tables. They
2806 default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
2807 you will need to set these values.
2808
2809 57. Updated the building process for Windows DLL, as provided by Fred Cox.
2810
2811
2812 Version 3.9 02-Jan-02
2813 ---------------------
2814
2815 1. A bit of extraneous text had somehow crept into the pcregrep documentation.
2816
2817 2. If --disable-static was given, the building process failed when trying to
2818 build pcretest and pcregrep. (For some reason it was using libtool to compile
2819 them, which is not right, as they aren't part of the library.)
2820
2821
2822 Version 3.8 18-Dec-01
2823 ---------------------
2824
2825 1. The experimental UTF-8 code was completely screwed up. It was packing the
2826 bytes in the wrong order. How dumb can you get?
2827
2828
2829 Version 3.7 29-Oct-01
2830 ---------------------
2831
2832 1. In updating pcretest to check change 1 of version 3.6, I screwed up.
2833 This caused pcretest, when used on the test data, to segfault. Unfortunately,
2834 this didn't happen under Solaris 8, where I normally test things.
2835
2836 2. The Makefile had to be changed to make it work on BSD systems, where 'make'
2837 doesn't seem to recognize that ./xxx and xxx are the same file. (This entry
2838 isn't in ChangeLog distributed with 3.7 because I forgot when I hastily made
2839 this fix an hour or so after the initial 3.7 release.)
2840
2841
2842 Version 3.6 23-Oct-01
2843 ---------------------
2844
2845 1. Crashed with /(sens|respons)e and \1ibility/ and "sense and sensibility" if
2846 offsets passed as NULL with zero offset count.
2847
2848 2. The config.guess and config.sub files had not been updated when I moved to
2849 the latest autoconf.
2850
2851
2852 Version 3.5 15-Aug-01
2853 ---------------------
2854
2855 1. Added some missing #if !defined NOPOSIX conditionals in pcretest.c that
2856 had been forgotten.
2857
2858 2. By using declared but undefined structures, we can avoid using "void"
2859 definitions in pcre.h while keeping the internal definitions of the structures
2860 private.
2861
2862 3. The distribution is now built using autoconf 2.50 and libtool 1.4. From a
2863 user point of view, this means that both static and shared libraries are built
2864 by default, but this can be individually controlled. More of the work of
2865 handling this static/shared cases is now inside libtool instead of PCRE's make
2866 file.
2867
2868 4. The pcretest utility is now installed along with pcregrep because it is
2869 useful for users (to test regexs) and by doing this, it automatically gets
2870 relinked by libtool. The documentation has been turned into a man page, so
2871 there are now .1, .txt, and .html versions in /doc.
2872
2873 5. Upgrades to pcregrep:
2874 (i) Added long-form option names like gnu grep.
2875 (ii) Added --help to list all options with an explanatory phrase.
2876 (iii) Added -r, --recursive to recurse into sub-directories.
2877 (iv) Added -f, --file to read patterns from a file.
2878
2879 6. pcre_exec() was referring to its "code" argument before testing that
2880 argument for NULL (and giving an error if it was NULL).
2881
2882 7. Upgraded Makefile.in to allow for compiling in a different directory from
2883 the source directory.
2884
2885 8. Tiny buglet in pcretest: when pcre_fullinfo() was called to retrieve the
2886 options bits, the pointer it was passed was to an int instead of to an unsigned
2887 long int. This mattered only on 64-bit systems.
2888
2889 9. Fixed typo (3.4/1) in pcre.h again. Sigh. I had changed pcre.h (which is
2890 generated) instead of pcre.in, which it its source. Also made the same change
2891 in several of the .c files.
2892
2893 10. A new release of gcc defines printf() as a macro, which broke pcretest
2894 because it had an ifdef in the middle of a string argument for printf(). Fixed
2895 by using separate calls to printf().
2896
2897 11. Added --enable-newline-is-cr and --enable-newline-is-lf to the configure
2898 script, to force use of CR or LF instead of \n in the source. On non-Unix
2899 systems, the value can be set in config.h.
2900
2901 12. The limit of 200 on non-capturing parentheses is a _nesting_ limit, not an
2902 absolute limit. Changed the text of the error message to make this clear, and
2903 likewise updated the man page.
2904
2905 13. The limit of 99 on the number of capturing subpatterns has been removed.
2906 The new limit is 65535, which I hope will not be a "real" limit.
2907
2908
2909 Version 3.4 22-Aug-00
2910 ---------------------
2911
2912 1. Fixed typo in pcre.h: unsigned const char * changed to const unsigned char *.
2913
2914 2. Diagnose condition (?(0) as an error instead of crashing on matching.
2915
2916
2917 Version 3.3 01-Aug-00
2918 ---------------------
2919
2920 1. If an octal character was given, but the value was greater than \377, it
2921 was not getting masked to the least significant bits, as documented. This could
2922 lead to crashes in some systems.
2923
2924 2. Perl 5.6 (if not earlier versions) accepts classes like [a-\d] and treats
2925 the hyphen as a literal. PCRE used to give an error; it now behaves like Perl.
2926
2927 3. Added the functions pcre_free_substring() and pcre_free_substring_list().
2928 These just pass their arguments on to (pcre_free)(), but they are provided
2929 because some uses of PCRE bind it to non-C systems that can call its functions,
2930 but cannot call free() or pcre_free() directly.
2931
2932 4. Add "make test" as a synonym for "make check". Corrected some comments in
2933 the Makefile.
2934
2935 5. Add $(DESTDIR)/ in front of all the paths in the "install" target in the
2936 Makefile.
2937
2938 6. Changed the name of pgrep to pcregrep, because Solaris has introduced a
2939 command called pgrep for grepping around the active processes.
2940
2941 7. Added the beginnings of support for UTF-8 character strings.
2942
2943 8. Arranged for the Makefile to pass over the settings of CC, CFLAGS, and
2944 RANLIB to ./ltconfig so that they are used by libtool. I think these are all
2945 the relevant ones. (AR is not passed because ./ltconfig does its own figuring
2946 out for the ar command.)
2947
2948
2949 Version 3.2 12-May-00
2950 ---------------------
2951
2952 This is purely a bug fixing release.
2953
2954 1. If the pattern /((Z)+|A)*/ was matched agained ZABCDEFG it matched Z instead
2955 of ZA. This was just one example of several cases that could provoke this bug,
2956 which was introduced by change 9 of version 2.00. The code for breaking
2957 infinite loops after an iteration that matches an empty string was't working
2958 correctly.
2959
2960 2. The pcretest program was not imitating Perl correctly for the pattern /a*/g
2961 when matched against abbab (for example). After matching an empty string, it
2962 wasn't forcing anchoring when setting PCRE_NOTEMPTY for the next attempt; this
2963 caused it to match further down the string than it should.
2964
2965 3. The code contained an inclusion of sys/types.h. It isn't clear why this
2966 was there because it doesn't seem to be needed, and it causes trouble on some
2967 systems, as it is not a Standard C header. It has been removed.
2968
2969 4. Made 4 silly changes to the source to avoid stupid compiler warnings that
2970 were reported on the Macintosh. The changes were from
2971
2972 while ((c = *(++ptr)) != 0 && c != '\n');
2973 to
2974 while ((c = *(++ptr)) != 0 && c != '\n') ;
2975
2976 Totally extraordinary, but if that's what it takes...
2977
2978 5. PCRE is being used in one environment where neither memmove() nor bcopy() is
2979 available. Added HAVE_BCOPY and an autoconf test for it; if neither
2980 HAVE_MEMMOVE nor HAVE_BCOPY is set, use a built-in emulation function which
2981 assumes the way PCRE uses memmove() (always moving upwards).
2982
2983 6. PCRE is being used in one environment where strchr() is not available. There
2984 was only one use in pcre.c, and writing it out to avoid strchr() probably gives
2985 faster code anyway.
2986
2987
2988 Version 3.1 09-Feb-00
2989 ---------------------
2990
2991 The only change in this release is the fixing of some bugs in Makefile.in for
2992 the "install" target:
2993
2994 (1) It was failing to install pcreposix.h.
2995
2996 (2) It was overwriting the pcre.3 man page with the pcreposix.3 man page.
2997
2998
2999 Version 3.0 01-Feb-00
3000 ---------------------
3001
3002 1. Add support for the /+ modifier to perltest (to output $` like it does in
3003 pcretest).
3004
3005 2. Add support for the /g modifier to perltest.
3006
3007 3. Fix pcretest so that it behaves even more like Perl for /g when the pattern
3008 matches null strings.
3009
3010 4. Fix perltest so that it doesn't do unwanted things when fed an empty
3011 pattern. Perl treats empty patterns specially - it reuses the most recent
3012 pattern, which is not what we want. Replace // by /(?#)/ in order to avoid this
3013 effect.
3014
3015 5. The POSIX interface was broken in that it was just handing over the POSIX
3016 captured string vector to pcre_exec(), but (since release 2.00) PCRE has
3017 required a bigger vector, with some working space on the end. This means that
3018 the POSIX wrapper now has to get and free some memory, and copy the results.
3019
3020 6. Added some simple autoconf support, placing the test data and the
3021 documentation in separate directories, re-organizing some of the
3022 information files, and making it build pcre-config (a GNU standard). Also added
3023 libtool support for building PCRE as a shared library, which is now the
3024 default.
3025
3026 7. Got rid of the leading zero in the definition of PCRE_MINOR because 08 and
3027 09 are not valid octal constants. Single digits will be used for minor values
3028 less than 10.
3029
3030 8. Defined REG_EXTENDED and REG_NOSUB as zero in the POSIX header, so that
3031 existing programs that set these in the POSIX interface can use PCRE without
3032 modification.
3033
3034 9. Added a new function, pcre_fullinfo() with an extensible interface. It can
3035 return all that pcre_info() returns, plus additional data. The pcre_info()
3036 function is retained for compatibility, but is considered to be obsolete.
3037
3038 10. Added experimental recursion feature (?R) to handle one common case that
3039 Perl 5.6 will be able to do with (?p{...}).
3040
3041 11. Added support for POSIX character classes like [:alpha:], which Perl is
3042 adopting.
3043
3044
3045 Version 2.08 31-Aug-99
3046 ----------------------
3047
3048 1. When startoffset was not zero and the pattern began with ".*", PCRE was not
3049 trying to match at the startoffset position, but instead was moving forward to
3050 the next newline as if a previous match had failed.
3051
3052 2. pcretest was not making use of PCRE_NOTEMPTY when repeating for /g and /G,
3053 and could get into a loop if a null string was matched other than at the start
3054 of the subject.
3055
3056 3. Added definitions of PCRE_MAJOR and PCRE_MINOR to pcre.h so the version can
3057 be distinguished at compile time, and for completeness also added PCRE_DATE.
3058
3059 5. Added Paul Sokolovsky's minor changes to make it easy to compile a Win32 DLL
3060 in GnuWin32 environments.
3061
3062
3063 Version 2.07 29-Jul-99
3064 ----------------------
3065
3066 1. The documentation is now supplied in plain text form and HTML as well as in
3067 the form of man page sources.
3068
3069 2. C++ compilers don't like assigning (void *) values to other pointer types.
3070 In particular this affects malloc(). Although there is no problem in Standard
3071 C, I've put in casts to keep C++ compilers happy.
3072
3073 3. Typo on pcretest.c; a cast of (unsigned char *) in the POSIX regexec() call
3074 should be (const char *).
3075
3076 4. If NOPOSIX is defined, pcretest.c compiles without POSIX support. This may
3077 be useful for non-Unix systems who don't want to bother with the POSIX stuff.
3078 However, I haven't made this a standard facility. The documentation doesn't
3079 mention it, and the Makefile doesn't support it.
3080
3081 5. The Makefile now contains an "install" target, with editable destinations at
3082 the top of the file. The pcretest program is not installed.
3083
3084 6. pgrep -V now gives the PCRE version number and date.
3085
3086 7. Fixed bug: a zero repetition after a literal string (e.g. /abcde{0}/) was
3087 causing the entire string to be ignored, instead of just the last character.
3088
3089 8. If a pattern like /"([^\\"]+|\\.)*"/ is applied in the normal way to a
3090 non-matching string, it can take a very, very long time, even for strings of
3091 quite modest length, because of the nested recursion. PCRE now does better in
3092 some of these cases. It does this by remembering the last required literal
3093 character in the pattern, and pre-searching the subject to ensure it is present
3094 before running the real match. In other words, it applies a heuristic to detect
3095 some types of certain failure quickly, and in the above example, if presented
3096 with a string that has no trailing " it gives "no match" very quickly.
3097
3098 9. A new runtime option PCRE_NOTEMPTY causes null string matches to be ignored;
3099 other alternatives are tried instead.
3100
3101
3102 Version 2.06 09-Jun-99
3103 ----------------------
3104
3105 1. Change pcretest's output for amount of store used to show just the code
3106 space, because the remainder (the data block) varies in size between 32-bit and
3107 64-bit systems.
3108
3109 2. Added an extra argument to pcre_exec() to supply an offset in the subject to
3110 start matching at. This allows lookbehinds to work when searching for multiple
3111 occurrences in a string.
3112
3113 3. Added additional options to pcretest for testing multiple occurrences:
3114
3115 /+ outputs the rest of the string that follows a match
3116 /g loops for multiple occurrences, using the new startoffset argument
3117 /G loops for multiple occurrences by passing an incremented pointer
3118
3119 4. PCRE wasn't doing the "first character" optimization for patterns starting
3120 with \b or \B, though it was doing it for other lookbehind assertions. That is,
3121 it wasn't noticing that a match for a pattern such as /\bxyz/ has to start with
3122 the letter 'x'. On long subject strings, this gives a significant speed-up.
3123
3124
3125 Version 2.05 21-Apr-99
3126 ----------------------
3127
3128 1. Changed the type of magic_number from int to long int so that it works
3129 properly on 16-bit systems.
3130
3131 2. Fixed a bug which caused patterns starting with .* not to work correctly
3132 when the subject string contained newline characters. PCRE was assuming
3133 anchoring for such patterns in all cases, which is not correct because .* will
3134 not pass a newline unless PCRE_DOTALL is set. It now assumes anchoring only if
3135 DOTALL is set at top level; otherwise it knows that patterns starting with .*
3136 must be retried after every newline in the subject.
3137
3138
3139 Version 2.04 18-Feb-99
3140 ----------------------
3141
3142 1. For parenthesized subpatterns with repeats whose minimum was zero, the
3143 computation of the store needed to hold the pattern was incorrect (too large).
3144 If such patterns were nested a few deep, this could multiply and become a real
3145 problem.
3146
3147 2. Added /M option to pcretest to show the memory requirement of a specific
3148 pattern. Made -m a synonym of -s (which does this globally) for compatibility.
3149
3150 3. Subpatterns of the form (regex){n,m} (i.e. limited maximum) were being
3151 compiled in such a way that the backtracking after subsequent failure was
3152 pessimal. Something like (a){0,3} was compiled as (a)?(a)?(a)? instead of
3153 ((a)((a)(a)?)?)? with disastrous performance if the maximum was of any size.
3154
3155
3156 Version 2.03 02-Feb-99
3157 ----------------------
3158
3159 1. Fixed typo and small mistake in man page.
3160
3161 2. Added 4th condition (GPL supersedes if conflict) and created separate
3162 LICENCE file containing the conditions.
3163
3164 3. Updated pcretest so that patterns such as /abc\/def/ work like they do in
3165 Perl, that is the internal \ allows the delimiter to be included in the
3166 pattern. Locked out the use of \ as a delimiter. If \ immediately follows
3167 the final delimiter, add \ to the end of the pattern (to test the error).
3168
3169 4. Added the convenience functions for extracting substrings after a successful
3170 match. Updated pcretest to make it able to test these functions.
3171
3172
3173 Version 2.02 14-Jan-99
3174 ----------------------
3175
3176 1. Initialized the working variables associated with each extraction so that
3177 their saving and restoring doesn't refer to uninitialized store.
3178
3179 2. Put dummy code into study.c in order to trick the optimizer of the IBM C
3180 compiler for OS/2 into generating correct code. Apparently IBM isn't going to
3181 fix the problem.
3182
3183 3. Pcretest: the timing code wasn't using LOOPREPEAT for timing execution
3184 calls, and wasn't printing the correct value for compiling calls. Increased the
3185 default value of LOOPREPEAT, and the number of significant figures in the
3186 times.
3187
3188 4. Changed "/bin/rm" in the Makefile to "-rm" so it works on Windows NT.
3189
3190 5. Renamed "deftables" as "dftables" to get it down to 8 characters, to avoid
3191 a building problem on Windows NT with a FAT file system.
3192
3193
3194 Version 2.01 21-Oct-98
3195 ----------------------
3196
3197 1. Changed the API for pcre_compile() to allow for the provision of a pointer
3198 to character tables built by pcre_maketables() in the current locale. If NULL
3199 is passed, the default tables are used.
3200
3201
3202 Version 2.00 24-Sep-98
3203 ----------------------
3204
3205 1. Since the (>?) facility is in Perl 5.005, don't require PCRE_EXTRA to enable
3206 it any more.
3207
3208 2. Allow quantification of (?>) groups, and make it work correctly.
3209
3210 3. The first character computation wasn't working for (?>) groups.
3211
3212 4. Correct the implementation of \Z (it is permitted to match on the \n at the
3213 end of the subject) and add 5.005's \z, which really does match only at the
3214 very end of the subject.
3215
3216 5. Remove the \X "cut" facility; Perl doesn't have it, and (?> is neater.
3217
3218 6. Remove the ability to specify CASELESS, MULTILINE, DOTALL, and
3219 DOLLAR_END_ONLY at runtime, to make it possible to implement the Perl 5.005
3220 localized options. All options to pcre_study() were also removed.
3221
3222 7. Add other new features from 5.005:
3223
3224 $(?<= positive lookbehind
3225 $(?<! negative lookbehind
3226 (?imsx-imsx) added the unsetting capability
3227 such a setting is global if at outer level; local otherwise
3228 (?imsx-imsx:) non-capturing groups with option setting
3229 (?(cond)re|re) conditional pattern matching
3230
3231 A backreference to itself in a repeated group matches the previous
3232 captured string.
3233
3234 8. General tidying up of studying (both automatic and via "study")
3235 consequential on the addition of new assertions.
3236
3237 9. As in 5.005, unlimited repeated groups that could match an empty substring
3238 are no longer faulted at compile time. Instead, the loop is forcibly broken at
3239 runtime if any iteration does actually match an empty substring.
3240
3241 10. Include the RunTest script in the distribution.
3242
3243 11. Added tests from the Perl 5.005_02 distribution. This showed up a few
3244 discrepancies, some of which were old and were also with respect to 5.004. They
3245 have now been fixed.
3246
3247
3248 Version 1.09 28-Apr-98
3249 ----------------------
3250
3251 1. A negated single character class followed by a quantifier with a minimum
3252 value of one (e.g. [^x]{1,6} ) was not compiled correctly. This could lead to
3253 program crashes, or just wrong answers. This did not apply to negated classes
3254 containing more than one character, or to minima other than one.
3255
3256
3257 Version 1.08 27-Mar-98
3258 ----------------------
3259
3260 1. Add PCRE_UNGREEDY to invert the greediness of quantifiers.
3261
3262 2. Add (?U) and (?X) to set PCRE_UNGREEDY and PCRE_EXTRA respectively. The
3263 latter must appear before anything that relies on it in the pattern.
3264
3265
3266 Version 1.07 16-Feb-98
3267 ----------------------
3268
3269 1. A pattern such as /((a)*)*/ was not being diagnosed as in error (unlimited
3270 repeat of a potentially empty string).
3271
3272
3273 Version 1.06 23-Jan-98
3274 ----------------------
3275
3276 1. Added Markus Oberhumer's little patches for C++.
3277
3278 2. Literal strings longer than 255 characters were broken.
3279
3280
3281 Version 1.05 23-Dec-97
3282 ----------------------
3283
3284 1. Negated character classes containing more than one character were failing if
3285 PCRE_CASELESS was set at run time.
3286
3287
3288 Version 1.04 19-Dec-97
3289 ----------------------
3290
3291 1. Corrected the man page, where some "const" qualifiers had been omitted.
3292
3293 2. Made debugging output print "{0,xxx}" instead of just "{,xxx}" to agree with
3294 input syntax.
3295
3296 3. Fixed memory leak which occurred when a regex with back references was
3297 matched with an offsets vector that wasn't big enough. The temporary memory
3298 that is used in this case wasn't being freed if the match failed.
3299
3300 4. Tidied pcretest to ensure it frees memory that it gets.
3301
3302 5. Temporary memory was being obtained in the case where the passed offsets
3303 vector was exactly big enough.
3304
3305 6. Corrected definition of offsetof() from change 5 below.
3306
3307 7. I had screwed up change 6 below and broken the rules for the use of
3308 setjmp(). Now fixed.
3309
3310
3311 Version 1.03 18-Dec-97
3312 ----------------------
3313
3314 1. A erroneous regex with a missing opening parenthesis was correctly
3315 diagnosed, but PCRE attempted to access brastack[-1], which could cause crashes
3316 on some systems.
3317
3318 2. Replaced offsetof(real_pcre, code) by offsetof(real_pcre, code[0]) because
3319 it was reported that one broken compiler failed on the former because "code" is
3320 also an independent variable.
3321
3322 3. The erroneous regex a[]b caused an array overrun reference.
3323
3324 4. A regex ending with a one-character negative class (e.g. /[^k]$/) did not
3325 fail on data ending with that character. (It was going on too far, and checking
3326 the next character, typically a binary zero.) This was specific to the
3327 optimized code for single-character negative classes.
3328
3329 5. Added a contributed patch from the TIN world which does the following:
3330
3331 + Add an undef for memmove, in case the the system defines a macro for it.
3332
3333 + Add a definition of offsetof(), in case there isn't one. (I don't know
3334 the reason behind this - offsetof() is part of the ANSI standard - but
3335 it does no harm).
3336
3337 + Reduce the ifdef's in pcre.c using macro DPRINTF, thereby eliminating
3338 most of the places where whitespace preceded '#'. I have given up and
3339 allowed the remaining 2 cases to be at the margin.
3340
3341 + Rename some variables in pcre to eliminate shadowing. This seems very
3342 pedantic, but does no harm, of course.
3343
3344 6. Moved the call to setjmp() into its own function, to get rid of warnings
3345 from gcc -Wall, and avoided calling it at all unless PCRE_EXTRA is used.
3346
3347 7. Constructs such as \d{8,} were compiling into the equivalent of
3348 \d{8}\d{0,65527} instead of \d{8}\d* which didn't make much difference to the
3349 outcome, but in this particular case used more store than had been allocated,
3350 which caused the bug to be discovered because it threw up an internal error.
3351
3352 8. The debugging code in both pcre and pcretest for outputting the compiled
3353 form of a regex was going wrong in the case of back references followed by
3354 curly-bracketed repeats.
3355
3356
3357 Version 1.02 12-Dec-97
3358 ----------------------
3359
3360 1. Typos in pcre.3 and comments in the source fixed.
3361
3362 2. Applied a contributed patch to get rid of places where it used to remove
3363 'const' from variables, and fixed some signed/unsigned and uninitialized
3364 variable warnings.
3365
3366 3. Added the "runtest" target to Makefile.
3367
3368 4. Set default compiler flag to -O2 rather than just -O.
3369
3370
3371 Version 1.01 19-Nov-97
3372 ----------------------
3373
3374 1. PCRE was failing to diagnose unlimited repeat of empty string for patterns
3375 like /([ab]*)*/, that is, for classes with more than one character in them.
3376
3377 2. Likewise, it wasn't diagnosing patterns with "once-only" subpatterns, such
3378 as /((?>a*))*/ (a PCRE_EXTRA facility).
3379
3380
3381 Version 1.00 18-Nov-97
3382 ----------------------
3383
3384 1. Added compile-time macros to support systems such as SunOS4 which don't have
3385 memmove() or strerror() but have other things that can be used instead.
3386
3387 2. Arranged that "make clean" removes the executables.
3388
3389
3390 Version 0.99 27-Oct-97
3391 ----------------------
3392
3393 1. Fixed bug in code for optimizing classes with only one character. It was
3394 initializing a 32-byte map regardless, which could cause it to run off the end
3395 of the memory it had got.
3396
3397 2. Added, conditional on PCRE_EXTRA, the proposed (?>REGEX) construction.
3398
3399
3400 Version 0.98 22-Oct-97
3401 ----------------------
3402
3403 1. Fixed bug in code for handling temporary memory usage when there are more
3404 back references than supplied space in the ovector. This could cause segfaults.
3405
3406
3407 Version 0.97 21-Oct-97
3408 ----------------------
3409
3410 1. Added the \X "cut" facility, conditional on PCRE_EXTRA.
3411
3412 2. Optimized negated single characters not to use a bit map.
3413
3414 3. Brought error texts together as macro definitions; clarified some of them;
3415 fixed one that was wrong - it said "range out of order" when it meant "invalid
3416 escape sequence".
3417
3418 4. Changed some char * arguments to const char *.
3419
3420 5. Added PCRE_NOTBOL and PCRE_NOTEOL (from POSIX).
3421
3422 6. Added the POSIX-style API wrapper in pcreposix.a and testing facilities in
3423 pcretest.
3424
3425
3426 Version 0.96 16-Oct-97
3427 ----------------------
3428
3429 1. Added a simple "pgrep" utility to the distribution.
3430
3431 2. Fixed an incompatibility with Perl: "{" is now treated as a normal character
3432 unless it appears in one of the precise forms "{ddd}", "{ddd,}", or "{ddd,ddd}"
3433 where "ddd" means "one or more decimal digits".
3434
3435 3. Fixed serious bug. If a pattern had a back reference, but the call to
3436 pcre_exec() didn't supply a large enough ovector to record the related
3437 identifying subpattern, the match always failed. PCRE now remembers the number
3438 of the largest back reference, and gets some temporary memory in which to save
3439 the offsets during matching if necessary, in order to ensure that
3440 backreferences always work.
3441
3442 4. Increased the compatibility with Perl in a number of ways:
3443
3444 (a) . no longer matches \n by default; an option PCRE_DOTALL is provided
3445 to request this handling. The option can be set at compile or exec time.
3446
3447 (b) $ matches before a terminating newline by default; an option
3448 PCRE_DOLLAR_ENDONLY is provided to override this (but not in multiline
3449 mode). The option can be set at compile or exec time.
3450
3451 (c) The handling of \ followed by a digit other than 0 is now supposed to be
3452 the same as Perl's. If the decimal number it represents is less than 10
3453 or there aren't that many previous left capturing parentheses, an octal
3454 escape is read. Inside a character class, it's always an octal escape,
3455 even if it is a single digit.
3456
3457 (d) An escaped but undefined alphabetic character is taken as a literal,
3458 unless PCRE_EXTRA is set. Currently this just reserves the remaining
3459 escapes.
3460
3461 (e) {0} is now permitted. (The previous item is removed from the compiled
3462 pattern).
3463
3464 5. Changed all the names of code files so that the basic parts are no longer
3465 than 10 characters, and abolished the teeny "globals.c" file.
3466
3467 6. Changed the handling of character classes; they are now done with a 32-byte
3468 bit map always.
3469
3470 7. Added the -d and /D options to pcretest to make it possible to look at the
3471 internals of compilation without having to recompile pcre.
3472
3473
3474 Version 0.95 23-Sep-97
3475 ----------------------
3476
3477 1. Fixed bug in pre-pass concerning escaped "normal" characters such as \x5c or
3478 \x20 at the start of a run of normal characters. These were being treated as
3479 real characters, instead of the source characters being re-checked.
3480
3481
3482 Version 0.94 18-Sep-97
3483 ----------------------
3484
3485 1. The functions are now thread-safe, with the caveat that the global variables
3486 containing pointers to malloc() and free() or alternative functions are the
3487 same for all threads.
3488
3489 2. Get pcre_study() to generate a bitmap of initial characters for non-
3490 anchored patterns when this is possible, and use it if passed to pcre_exec().
3491
3492
3493 Version 0.93 15-Sep-97
3494 ----------------------
3495
3496 1. /(b)|(:+)/ was computing an incorrect first character.
3497
3498 2. Add pcre_study() to the API and the passing of pcre_extra to pcre_exec(),
3499 but not actually doing anything yet.
3500
3501 3. Treat "-" characters in classes that cannot be part of ranges as literals,
3502 as Perl does (e.g. [-az] or [az-]).
3503
3504 4. Set the anchored flag if a branch starts with .* or .*? because that tests
3505 all possible positions.
3506
3507 5. Split up into different modules to avoid including unneeded functions in a
3508 compiled binary. However, compile and exec are still in one module. The "study"
3509 function is split off.
3510
3511 6. The character tables are now in a separate module whose source is generated
3512 by an auxiliary program - but can then be edited by hand if required. There are
3513 now no calls to isalnum(), isspace(), isdigit(), isxdigit(), tolower() or
3514 toupper() in the code.
3515
3516 7. Turn the malloc/free funtions variables into pcre_malloc and pcre_free and
3517 make them global. Abolish the function for setting them, as the caller can now
3518 set them directly.
3519
3520
3521 Version 0.92 11-Sep-97
3522 ----------------------
3523
3524 1. A repeat with a fixed maximum and a minimum of 1 for an ordinary character
3525 (e.g. /a{1,3}/) was broken (I mis-optimized it).
3526
3527 2. Caseless matching was not working in character classes if the characters in
3528 the pattern were in upper case.
3529
3530 3. Make ranges like [W-c] work in the same way as Perl for caseless matching.
3531
3532 4. Make PCRE_ANCHORED public and accept as a compile option.
3533
3534 5. Add an options word to pcre_exec() and accept PCRE_ANCHORED and
3535 PCRE_CASELESS at run time. Add escapes \A and \I to pcretest to cause it to
3536 pass them.
3537
3538 6. Give an error if bad option bits passed at compile or run time.
3539
3540 7. Add PCRE_MULTILINE at compile and exec time, and (?m) as well. Add \M to
3541 pcretest to cause it to pass that flag.
3542
3543 8. Add pcre_info(), to get the number of identifying subpatterns, the stored
3544 options, and the first character, if set.
3545
3546 9. Recognize C+ or C{n,m} where n >= 1 as providing a fixed starting character.
3547
3548
3549 Version 0.91 10-Sep-97
3550 ----------------------
3551
3552 1. PCRE was failing to diagnose unlimited repeats of subpatterns that could
3553 match the empty string as in /(a*)*/. It was looping and ultimately crashing.
3554
3555 2. PCRE was looping on encountering an indefinitely repeated back reference to
3556 a subpattern that had matched an empty string, e.g. /(a|)\1*/. It now does what
3557 Perl does - treats the match as successful.
3558
3559 ****

Properties

Name Value
svn:eol-style native
svn:keywords "Author Date Id Revision Url"

webmaster@exim.org
ViewVC Help
Powered by ViewVC 1.1.12